lib-x17-fintech 2.1.3__py3-none-any.whl

xfintech/data/source/tushare/stock/stockdividend/tests/test_class_stockdividend_all.py

@@ -0,0 +1,725 @@
+"""
+Comprehensive test suite for StockDividend class
+Tests cover initialization, transformation, execution, and integration scenarios
+"""
+
+import pandas as pd
+import pytest
+
+from xfintech.data.source.tushare.stock.stockdividend import StockDividend
+from xfintech.data.source.tushare.stock.stockdividend.constant import (
+    KEY,
+    NAME,
+    PAGINATE,
+    SOURCE,
+    TARGET,
+)
+
+# ============================================================================
+# Test Fixtures
+# ============================================================================
+
+
+class FakeConnection:
+    """Fake Tushare connection for testing"""
+
+    def __init__(self, frame: pd.DataFrame):
+        self.frame = frame
+
+    def dividend(self, **kwargs):
+        """Mock dividend API call"""
+        return self.frame
+
+
+class FakeSession:
+    """Fake session for testing"""
+
+    def __init__(self, connection: FakeConnection):
+        self.connection = connection
+
+
+@pytest.fixture
+def mock_session():
+    """Create a mock session with empty data"""
+    fake_conn = FakeConnection(frame=pd.DataFrame())
+    return FakeSession(fake_conn)
+
+
+@pytest.fixture
+def sample_source_data():
+    """Sample dividend data in Tushare format"""
+    return pd.DataFrame(
+        {
+            "ts_code": ["600848.SH", "600848.SH", "000001.SZ"],
+            "end_date": ["20201231", "20191231", "20201231"],
+            "ann_date": ["20210415", "20200410", "20210420"],
+            "div_proc": ["实施", "实施", "预案"],
+            "stk_div": ["0.10", "0.15", "0.00"],
+            "stk_bo_rate": ["0.05", "0.08", "0.00"],
+            "stk_co_rate": ["0.05", "0.07", "0.00"],
+            "cash_div": ["0.50", "0.60", "0.30"],
+            "cash_div_tax": ["0.55", "0.65", "0.32"],
+            "record_date": ["20210520", "20200515", "20210525"],
+            "ex_date": ["20210521", "20200516", "20210526"],
+            "pay_date": ["20210528", "20200525", "20210601"],
+            "div_listdate": ["20210528", "20200525", "20210601"],
+            "imp_ann_date": ["20210415", "20200410", "20210420"],
+            "base_date": ["20201231", "20191231", "20201231"],
+            "base_share": ["100000.00", "95000.00", "200000.00"],
+        }
+    )
+
+
+# ============================================================================
+# Initialization Tests
+# ============================================================================
+
+
+def test_stockdividend_init_basic(mock_session):
+    """Test basic initialization with required session"""
+    dividend = StockDividend(session=mock_session)
+    assert dividend.name == NAME
+    assert dividend.key == KEY
+    assert dividend.source == SOURCE
+    assert dividend.target == TARGET
+    assert dividend.paginate == PAGINATE
+
+
+def test_stockdividend_init_with_params_dict(mock_session):
+    """Test initialization with params as dict"""
+    params = {"ts_code": "600848.SH", "ann_date": "20210415"}
+    dividend = StockDividend(session=mock_session, params=params)
+    assert dividend.params.get("ts_code") == "600848.SH"
+    assert dividend.params.get("ann_date") == "20210415"
+
+
+def test_stockdividend_init_with_params_object(mock_session):
+    """Test initialization with params as Params object"""
+    from xfintech.data.common.params import Params
+
+    params = Params(ts_code="600848.SH")
+    dividend = StockDividend(session=mock_session, params=params)
+    assert dividend.params.get("ts_code") == "600848.SH"
+
+
+def test_stockdividend_init_with_multiple_date_params(mock_session):
+    """Test initialization with multiple date parameters"""
+    params = {
+        "ts_code": "600848.SH",
+        "ann_date": "20210415",
+        "record_date": "20210520",
+        "ex_date": "20210521",
+    }
+    dividend = StockDividend(session=mock_session, params=params)
+    assert dividend.params.get("ann_date") == "20210415"
+    assert dividend.params.get("record_date") == "20210520"
+
+
+def test_stockdividend_init_with_cache_bool_true(mock_session):
+    """Test initialization with cache enabled"""
+    dividend = StockDividend(session=mock_session, cache=True)
+    assert dividend.cache is not None
+
+
+def test_stockdividend_init_with_cache_bool_false(mock_session):
+    """Test initialization with cache disabled"""
+    dividend = StockDividend(session=mock_session, cache=False)
+    assert dividend.cache is None
+
+
+def test_stockdividend_init_with_cache_dict(mock_session):
+    """Test initialization with cache config dict"""
+    cache_config = {"path": "/tmp/cache", "key": "test"}
+    dividend = StockDividend(session=mock_session, cache=cache_config)
+    assert dividend.cache is not None
+
+
+def test_stockdividend_init_default_paginate_limit(mock_session):
+    """Test that default pagination limit is set correctly"""
+    dividend = StockDividend(session=mock_session)
+    assert dividend.paginate.pagesize == 1000
+    assert dividend.paginate.pagelimit == 1000
+
+
+def test_stockdividend_init_with_all_params(mock_session):
+    """Test initialization with all parameters"""
+    params = {"ts_code": "600848.SH", "ann_date": "20210415"}
+    coolant = {"interval": 0.5}
+    retry = {"max_attempts": 3}
+    cache = True
+
+    dividend = StockDividend(session=mock_session, params=params, coolant=coolant, retry=retry, cache=cache)
+    assert dividend.params.get("ts_code") == "600848.SH"
+    assert dividend.cache is not None
+
+
+def test_stockdividend_constants():
+    """Test that constants are correctly defined"""
+    assert NAME == "stockdividend"
+    assert KEY == "/tushare/stockdividend"
+    assert SOURCE.name == "dividend"
+    assert TARGET.name == "stockdividend"
+    assert PAGINATE.pagesize == 1000
+    assert PAGINATE.pagelimit == 1000
+
+
+# ============================================================================
+# Transform Tests
+# ============================================================================
+
+
+def test_stockdividend_transform_basic(sample_source_data):
+    """Test basic transformation of dividend data"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session)
+
+    result = dividend.transform(sample_source_data)
+
+    assert len(result) == 3
+    assert "code" in result.columns
+    assert "ann_date" in result.columns
+    assert "ann_datecode" in result.columns
+
+
+def test_stockdividend_transform_code_mapping(sample_source_data):
+    """Test that ts_code is correctly mapped to code"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session)
+
+    result = dividend.transform(sample_source_data)
+
+    assert result.iloc[0]["code"] == "000001.SZ"
+    assert result.iloc[1]["code"] == "600848.SH"
+
+
+def test_stockdividend_transform_date_format(sample_source_data):
+    """Test that dates are properly formatted"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session)
+
+    result = dividend.transform(sample_source_data)
+
+    assert pd.api.types.is_datetime64_any_dtype(result["ann_date"]) is False
+    assert result.iloc[1]["ann_date"] == "2020-04-10"
+    assert result.iloc[1]["record_date"] == "2020-05-15"
+
+
+def test_stockdividend_transform_datecode_preserved(sample_source_data):
+    """Test that datecode is correctly created"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session)
+
+    result = dividend.transform(sample_source_data)
+
+    assert result.iloc[1]["ann_datecode"] == "20200410"
+    assert result.iloc[1]["record_datecode"] == "20200515"
+    assert result.iloc[1]["ex_datecode"] == "20200516"
+
+
+def test_stockdividend_transform_numeric_conversions(sample_source_data):
+    """Test that numeric fields are properly converted"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session)
+
+    result = dividend.transform(sample_source_data)
+
+    assert result.iloc[1]["stk_div"] == 0.15
+    assert result.iloc[1]["cash_div"] == 0.60
+    assert result.iloc[1]["base_share"] == 95000.00
+    assert pd.api.types.is_numeric_dtype(result["stk_div"])
+
+
+def test_stockdividend_transform_string_fields(sample_source_data):
+    """Test that string fields are properly converted"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session)
+
+    result = dividend.transform(sample_source_data)
+
+    assert result.iloc[1]["div_proc"] == "实施"
+    assert result.iloc[0]["div_proc"] == "预案"
+    assert result.iloc[1]["end_date"] == "20191231"
+
+
+def test_stockdividend_transform_empty_dataframe():
+    """Test transformation with empty dataframe"""
+    fake_conn = FakeConnection(frame=pd.DataFrame())
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session)
+
+    result = dividend.transform(pd.DataFrame())
+
+    assert isinstance(result, pd.DataFrame)
+    assert result.empty
+    assert len(result.columns) == len(TARGET.list_column_names())
+
+
+def test_stockdividend_transform_none_input():
+    """Test transformation with None input"""
+    fake_conn = FakeConnection(frame=pd.DataFrame())
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session)
+
+    result = dividend.transform(None)
+
+    assert isinstance(result, pd.DataFrame)
+    assert result.empty
+
+
+def test_stockdividend_transform_handles_invalid_dates():
+    """Test that invalid dates are handled gracefully"""
+    data = pd.DataFrame(
+        {
+            "ts_code": ["600848.SH"],
+            "end_date": ["20201231"],
+            "ann_date": ["invalid"],
+            "div_proc": ["实施"],
+            "stk_div": ["0.10"],
+            "cash_div": ["0.50"],
+            "record_date": ["20210520"],
+            "ex_date": ["20210521"],
+            "pay_date": ["20210528"],
+            "div_listdate": ["20210528"],
+            "imp_ann_date": ["20210415"],
+            "base_date": ["20201231"],
+        }
+    )
+
+    fake_conn = FakeConnection(frame=data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session)
+
+    result = dividend.transform(data)
+
+    assert len(result) == 1
+    assert pd.isna(result.iloc[0]["ann_date"])
+
+
+def test_stockdividend_transform_removes_duplicates(sample_source_data):
+    """Test that duplicate rows are removed"""
+    # Create data with duplicates
+    data = pd.concat([sample_source_data, sample_source_data.iloc[[0]]], ignore_index=True)
+
+    fake_conn = FakeConnection(frame=data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session)
+
+    result = dividend.transform(data)
+
+    # Should have 3 unique rows, not 4
+    assert len(result) == 3
+
+
+def test_stockdividend_transform_sorts_by_code_and_ann_date(sample_source_data):
+    """Test that output is sorted by code and ann_date"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session)
+
+    result = dividend.transform(sample_source_data)
+
+    # Check sorting: 000001.SZ first, then 600848.SH by dates
+    assert result.iloc[0]["code"] == "000001.SZ"
+    assert result.iloc[1]["code"] == "600848.SH"
+    assert result.iloc[1]["ann_date"] == "2020-04-10"
+    assert result.iloc[2]["code"] == "600848.SH"
+    assert result.iloc[2]["ann_date"] == "2021-04-15"
+
+
+def test_stockdividend_transform_resets_index():
+    """Test that index is reset after transformation"""
+    data = pd.DataFrame(
+        {
+            "ts_code": ["600848.SH", "000001.SZ"],
+            "end_date": ["20201231", "20201231"],
+            "ann_date": ["20210415", "20210420"],
+            "div_proc": ["实施", "预案"],
+            "stk_div": ["0.10", "0.00"],
+            "cash_div": ["0.50", "0.30"],
+            "record_date": ["20210520", "20210525"],
+            "ex_date": ["20210521", "20210526"],
+            "pay_date": ["20210528", "20210601"],
+            "div_listdate": ["20210528", "20210601"],
+            "imp_ann_date": ["20210415", "20210420"],
+            "base_date": ["20201231", "20201231"],
+        },
+        index=[5, 10],
+    )
+
+    fake_conn = FakeConnection(frame=data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session)
+
+    result = dividend.transform(data)
+
+    assert result.index.tolist() == [0, 1]
+
+
+def test_stockdividend_transform_only_target_columns(sample_source_data):
+    """Test that only target columns are in output"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session)
+
+    result = dividend.transform(sample_source_data)
+
+    expected_cols = set(TARGET.list_column_names())
+    actual_cols = set(result.columns)
+
+    assert actual_cols == expected_cols
+
+
+def test_stockdividend_transform_all_date_fields(sample_source_data):
+    """Test that all date fields are properly transformed"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session)
+
+    result = dividend.transform(sample_source_data)
+
+    # Check that all date/datecode pairs exist
+    date_pairs = [
+        ("ann_date", "ann_datecode"),
+        ("record_date", "record_datecode"),
+        ("ex_date", "ex_datecode"),
+        ("pay_date", "pay_datecode"),
+        ("div_listdate", "div_listdatecode"),
+        ("imp_ann_date", "imp_ann_datecode"),
+        ("base_date", "base_datecode"),
+    ]
+
+    for date_col, datecode_col in date_pairs:
+        assert date_col in result.columns
+        assert datecode_col in result.columns
+
+
+# ============================================================================
+# Run Method Tests
+# ============================================================================
+
+
+def test_stockdividend_run_basic(sample_source_data):
+    """Test basic run() execution"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session, params={"ts_code": "600848.SH"})
+
+    result = dividend.run()
+
+    assert isinstance(result, pd.DataFrame)
+    assert len(result) > 0
+
+
+def test_stockdividend_run_with_cache_hit(sample_source_data):
+    """Test that run() returns cached data when available"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session, params={"ts_code": "600848.SH"}, cache=True)
+
+    # First run should fetch and cache
+    result1 = dividend.run()
+
+    # Second run should return cached data
+    result2 = dividend.run()
+
+    pd.testing.assert_frame_equal(result1, result2)
+
+
+def test_stockdividend_run_with_ann_date(sample_source_data):
+    """Test run() with ann_date parameter"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session, params={"ann_date": "20210415"})
+
+    result = dividend.run()
+
+    assert isinstance(result, pd.DataFrame)
+
+
+def test_stockdividend_run_with_datetime_params(sample_source_data):
+    """Test run() with datetime objects as parameters"""
+    from datetime import datetime
+
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(
+        session=session,
+        params={
+            "ts_code": "600848.SH",
+            "ann_date": datetime(2021, 4, 15),
+        },
+    )
+
+    result = dividend.run()
+
+    assert isinstance(result, pd.DataFrame)
+
+
+def test_stockdividend_run_adds_fields_param(sample_source_data):
+    """Test that run() adds fields parameter if not provided"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session, params={"ts_code": "600848.SH"})
+
+    # Monkey patch _fetchall to capture params
+    dividend._fetchall
+    captured_params = {}
+
+    def mock_fetchall(api, **kwargs):
+        captured_params.update(kwargs)
+        return sample_source_data
+
+    dividend._fetchall = mock_fetchall
+    dividend.run()
+
+    assert "fields" in captured_params
+    assert len(captured_params["fields"]) > 0
+
+
+def test_stockdividend_run_sets_cache(sample_source_data):
+    """Test that run() saves result to cache"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session, params={"ts_code": "600848.SH"}, cache=True)
+
+    result = dividend.run()
+
+    # Cache should be set after run
+    cached = dividend._load_cache()
+    assert cached is not None
+    pd.testing.assert_frame_equal(result, cached)
+
+
+def test_stockdividend_run_calls_transform(sample_source_data):
+    """Test that run() calls transform method"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session, params={"ts_code": "600848.SH"})
+
+    # Monkey patch transform to track if it was called
+    original_transform = dividend.transform
+    transform_called = {"called": False}
+
+    def mock_transform(data):
+        transform_called["called"] = True
+        return original_transform(data)
+
+    dividend.transform = mock_transform
+    dividend.run()
+
+    assert transform_called["called"] is True
+
+
+def test_stockdividend_run_uses_dividend_api(sample_source_data):
+    """Test that run() uses dividend API"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session, params={"ts_code": "600848.SH"})
+
+    # Monkey patch _fetchall to capture API
+    captured_api = {"api": None}
+
+    def mock_fetchall(api, **kwargs):
+        captured_api["api"] = api
+        return sample_source_data
+
+    dividend._fetchall = mock_fetchall
+    dividend.run()
+
+    assert captured_api["api"] == session.connection.dividend
+
+
+# ============================================================================
+# Integration Tests
+# ============================================================================
+
+
+def test_stockdividend_full_workflow(sample_source_data):
+    """Test complete workflow from initialization to output"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+
+    dividend = StockDividend(
+        session=session,
+        params={"ts_code": "600848.SH"},
+        cache=True,
+    )
+
+    result = dividend.run()
+
+    # Verify output structure
+    assert isinstance(result, pd.DataFrame)
+    assert len(result) == 3
+    assert "code" in result.columns
+    assert "ann_date" in result.columns
+    assert "div_proc" in result.columns
+    assert "stk_div" in result.columns
+
+    # Verify data types
+    assert pd.api.types.is_numeric_dtype(result["stk_div"])
+    assert pd.api.types.is_numeric_dtype(result["cash_div"])
+
+
+def test_stockdividend_cache_persistence(sample_source_data):
+    """Test that cache persists across multiple runs"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session, params={"ts_code": "600848.SH"}, cache=True)
+
+    # Run once to populate cache
+    result1 = dividend.run()
+
+    # Clear in-memory data to simulate fresh load
+    dividend2 = StockDividend(session=session, params={"ts_code": "600848.SH"}, cache=True)
+    result2 = dividend2.run()
+
+    pd.testing.assert_frame_equal(result1, result2)
+
+
+def test_stockdividend_params_uniqueness(mock_session):
+    """Test that different params produce different cache keys"""
+    dividend1 = StockDividend(
+        session=mock_session,
+        params={"ts_code": "600848.SH"},
+        cache=True,
+    )
+    dividend2 = StockDividend(
+        session=mock_session,
+        params={"ts_code": "000001.SZ"},
+        cache=True,
+    )
+
+    # Test that different params are properly stored
+    assert dividend1.params.get("ts_code") == "600848.SH"
+    assert dividend2.params.get("ts_code") == "000001.SZ"
+    assert dividend1.params.get("ts_code") != dividend2.params.get("ts_code")
+
+
+def test_stockdividend_empty_result_handling():
+    """Test handling of empty API results"""
+    empty_data = pd.DataFrame()
+    fake_conn = FakeConnection(frame=empty_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session, params={"ts_code": "600848.SH"})
+
+    result = dividend.run()
+
+    assert isinstance(result, pd.DataFrame)
+    assert result.empty
+    assert len(result.columns) > 0
+
+
+def test_stockdividend_large_dataset_handling():
+    """Test handling of large datasets"""
+    # Create large dataset
+    large_data = pd.DataFrame(
+        {
+            "ts_code": ["600848.SH"] * 1000,
+            "end_date": ["20201231"] * 1000,
+            "ann_date": ["20210415"] * 1000,
+            "div_proc": ["实施"] * 1000,
+            "stk_div": ["0.10"] * 1000,
+            "cash_div": ["0.50"] * 1000,
+            "record_date": ["20210520"] * 1000,
+            "ex_date": ["20210521"] * 1000,
+            "pay_date": ["20210528"] * 1000,
+            "div_listdate": ["20210528"] * 1000,
+            "imp_ann_date": ["20210415"] * 1000,
+            "base_date": ["20201231"] * 1000,
+        }
+    )
+
+    fake_conn = FakeConnection(frame=large_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session, params={"ts_code": "600848.SH"})
+
+    result = dividend.run()
+
+    # Should deduplicate to 1 row
+    assert len(result) == 1
+
+
+def test_stockdividend_without_cache(sample_source_data):
+    """Test that class works correctly without caching"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session, params={"ts_code": "600848.SH"}, cache=False)
+
+    result = dividend.run()
+
+    assert isinstance(result, pd.DataFrame)
+    assert len(result) > 0
+    assert dividend.cache is None
+
+
+def test_stockdividend_handles_missing_fields():
+    """Test that missing fields don't break transformation"""
+    # Data with only subset of fields
+    minimal_data = pd.DataFrame(
+        {
+            "ts_code": ["600848.SH"],
+            "end_date": ["20201231"],
+            "ann_date": ["20210415"],
+            "div_proc": ["实施"],
+            "stk_div": ["0.10"],
+            "cash_div": ["0.50"],
+            "record_date": ["20210520"],
+            "ex_date": ["20210521"],
+        }
+    )
+
+    fake_conn = FakeConnection(frame=minimal_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session, params={"ts_code": "600848.SH"})
+
+    result = dividend.run()
+
+    # Should complete without error
+    assert len(result) == 1
+    # Missing fields should be NaN
+    assert pd.isna(result.iloc[0]["base_share"])
+
+
+def test_stockdividend_api_doc_reference():
+    """Test that class correctly references API documentation"""
+    assert "doc_id=103" in StockDividend.__doc__
+
+
+def test_stockdividend_crosssection_scale():
+    """Test that module is correctly configured as crosssection scale"""
+    assert SOURCE.meta["scale"] == "crosssection"
+    assert TARGET.meta["scale"] == "crosssection"
+
+
+def test_stockdividend_multiple_stocks_handling(sample_source_data):
+    """Test handling of data with multiple stocks"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session, params={"ann_date": "20210415"})
+
+    result = dividend.run()
+
+    # Should have data for both stocks
+    assert len(result) == 3
+    assert "600848.SH" in result["code"].values
+    assert "000001.SZ" in result["code"].values
+
+
+def test_stockdividend_handles_zero_values(sample_source_data):
+    """Test that zero values in numeric fields are handled correctly"""
+    fake_conn = FakeConnection(frame=sample_source_data)
+    session = FakeSession(fake_conn)
+    dividend = StockDividend(session=session)
+
+    result = dividend.transform(sample_source_data)
+
+    # Check that zero values are preserved (not treated as missing)
+    zero_div_rows = result[result["stk_div"] == 0.0]
+    assert len(zero_div_rows) == 1
+    assert zero_div_rows.iloc[0]["code"] == "000001.SZ"