Skip to content

"Vibe-Trading: Your Personal Trading Agent"

6.3 KB
Raw Blame
"""Unit tests for optional Tushare fallback adapters.

The Tushare client is replaced by small in-memory fakes, so these tests never
touch the network or require a real token.
"""

from __future__ import annotations

import pytest

from types import SimpleNamespace
from unittest.mock import patch

from src.tools import tushare_fallbacks as tf


def test_fund_flow_maps_moneyflow_buckets_to_existing_schema() -> None:
    pro = SimpleNamespace(
        moneyflow=lambda **_: [
            {
                "trade_date": "20240103",
                "net_mf_amount": 12.5,
                "buy_sm_amount": 3.0,
                "sell_sm_amount": 1.0,
                "buy_md_amount": 5.0,
                "sell_md_amount": 8.0,
                "buy_lg_amount": 20.0,
                "sell_lg_amount": 7.0,
                "buy_elg_amount": 30.0,
                "sell_elg_amount": 10.0,
            }
        ]
    )
    with patch.object(tf, "_pro_api", return_value=pro), patch.object(
        tf, "_date_window", return_value=("20240101", "20240103")
    ):
        result = tf.fetch_fund_flow("600519.SH", days=5)

    row = result["rows"][0]
    assert result["source"] == "tushare"
    assert row["timestamp"] == "2024-01-03"
    assert row["main"] == 330000.0
    assert row["net_total"] == 125000.0
    assert row["small"] == 20000.0
    assert row["medium"] == -30000.0
    assert row["large"] == 130000.0
    assert row["super_large"] == 200000.0


def test_dragon_tiger_maps_top_list_and_top_inst() -> None:
    pro = SimpleNamespace(
        top_list=lambda **_: [
            {
                "ts_code": "600519.SH",
                "name": "Kweichow Moutai",
                "close": 1700.0,
                "pct_change": 5.2,
                "net_amount": 1.2e8,
                "l_buy": 3.0e8,
                "l_sell": 1.8e8,
                "amount": 9.0e8,
                "reason": "daily move",
            }
        ],
        top_inst=lambda **_: [
            {"exalter": "Institution", "side": "0", "buy": 2.0e8, "sell": 0.0, "net_buy": 2.0e8}
        ],
    )
    with patch.object(tf, "_pro_api", return_value=pro):
        data = tf.fetch_dragon_tiger("2024-01-02", "600519")

    assert data["date"] == "2024-01-02"
    assert data["appearances"][0]["code"] == "600519"
    assert data["appearances"][0]["net_buy"] == 1.2e8
    assert data["seats"][0]["seat"] == "Institution"
    assert data["seats"][0]["net"] == 2.0e8


def test_northbound_passes_through_million_yuan_with_semantics_note() -> None:
    """No ×100 rescale: post-2024-08-19 rows are turnover in million CNY
    (HKEX-verified, issue #1481); a constant factor cannot serve both eras,
    so values pass through and the note documents the boundary."""
    pro = SimpleNamespace(
        moneyflow_hsgt=lambda **_: [
            {"trade_date": "20240816", "hgt": -2568.22, "sgt": 100.0, "north_money": -2468.22},
            {"trade_date": "20240819", "hgt": 46129.34, "sgt": 50000.0, "north_money": 96129.34},
        ]
    )
    with patch.object(tf, "_pro_api", return_value=pro), patch.object(
        tf, "_date_window", return_value=("20240101", "20240103")
    ):
        data = tf.fetch_northbound_flow(lookback_days=2)

    assert data["unit"] == "CNY million"
    assert data["history"][0]["shanghai_connect"] == -2568.22
    assert data["history"][0]["total"] == -2468.22
    assert data["history"][0]["turnover_total"] is None
    assert data["history"][1]["shanghai_connect"] is None
    assert data["history"][1]["total"] is None
    assert data["history"][1]["turnover_shanghai"] == 46129.34
    assert data["history"][1]["turnover_total"] == 96129.34
    assert data["realtime"]["total"] is None
    assert data["realtime"]["turnover_total"] == 96129.34
    assert "2024-08-19" in data["note"]
    assert "turnover" in data["note"].lower()


def test_margin_trading_maps_and_sorts_most_recent_first() -> None:
    pro = SimpleNamespace(
        margin_detail=lambda **_: [
            {"trade_date": "20240102", "rzye": 1.0, "rzmre": 2.0, "rzche": 3.0, "rqye": 4.0, "rqyl": 5.0, "rzrqye": 6.0},
            {"trade_date": "20240103", "rzye": 7.0, "rzmre": 8.0, "rzche": 9.0, "rqye": 10.0, "rqyl": 11.0, "rzrqye": 12.0},
        ]
    )
    with patch.object(tf, "_pro_api", return_value=pro), patch.object(
        tf, "_date_window", return_value=("20240101", "20240103")
    ):
        data = tf.fetch_margin_trading("600519.SH", days=5)

    assert data["ts_code"] == "600519.SH"
    assert data["rows"][0]["trade_date"] == "2024-01-03"
    assert data["rows"][0]["financing_balance"] == 7.0
    assert data["rows"][1]["margin_total_balance"] == 6.0


def test_ts_code_routes_beijing_920_codes_to_bj() -> None:
    assert tf._ts_code("920002") == "920002.BJ"
    assert tf._ts_code("830799") == "830799.BJ"
    assert tf._ts_code("430047") == "430047.BJ"
    assert tf._ts_code("900901") == "900901.SH"
    assert tf._ts_code("600519") == "600519.SH"
    assert tf._ts_code("000001") == "000001.SZ"


@pytest.mark.parametrize("missing", [None, "-", float("nan"), float("inf")])
def test_fund_flow_missing_side_is_not_zero_or_total_flow(missing):
    pro = SimpleNamespace(moneyflow=lambda **_: [{
        "trade_date": "20240103", "net_mf_amount": 999,
        "buy_lg_amount": missing, "sell_lg_amount": 7,
        "buy_elg_amount": 30, "sell_elg_amount": 10,
    }])
    with patch.object(tf, "_pro_api", return_value=pro):
        row = tf.fetch_fund_flow("600519.SH", days=5)["rows"][0]
    assert row["large"] is None
    assert row["main"] is None
    assert row["small"] is None
    assert row["super_large"] == 200000


def test_empty_tushare_fund_flow_is_unavailable():
    with patch.object(tf, "_pro_api", return_value=SimpleNamespace(moneyflow=lambda **_: [])):
        with pytest.raises(tf.TushareFallbackUnavailable, match="no.*rows"):
            tf.fetch_fund_flow("600519.SH", days=5)


def test_real_zero_buckets_stay_zero_and_absent_total_stays_unknown():
    data = {"trade_date": "20240103"}
    for size in ("sm", "md", "lg", "elg"):
        data[f"buy_{size}_amount"] = data[f"sell_{size}_amount"] = 0
    with patch.object(tf, "_pro_api", return_value=SimpleNamespace(moneyflow=lambda **_: [data])):
        row = tf.fetch_fund_flow("600519.SH", days=5)["rows"][0]
    assert all(row[key] == 0 for key in ("main", "small", "medium", "large", "super_large"))
    assert row["net_total"] is None