import pandas as pd

from scripts.validate_accuracy_target_gate import (
    TargetThresholds,
    evaluate_accuracy_target,
    factor_names_for_configs,
)


def test_factor_names_union_keeps_candidate_order_then_production_extras():
    candidate = {"weights": {"high_base": 0.02, "short_term_battle": 0.03}, "hold_bias": 0.0}
    production = {"weights": {"oldwang_trend": 0.01, "high_base": 0.02}, "hold_bias": 0.0}

    assert factor_names_for_configs(candidate, production) == [
        "high_base",
        "short_term_battle",
        "oldwang_trend",
    ]


def test_accuracy_target_gate_passes_against_current_production_baseline():
    rows = []
    for idx in range(12):
        bullish = idx in {7, 8, 9}
        bearish = idx in {6, 10}
        rows.append(
            {
                "stock": "2330",
                "date": f"2026-05-{idx + 1:02d}",
                "y_true": 1 if bullish else (-1 if bearish else 0),
                "p_buy": 0.44,
                "p_hold": 0.48,
                "p_sell": 0.08,
                "base_pred": 0,
                "factor__new_signal": 1.0 if bullish else (-1.0 if bearish else 0.0),
                "factor__old_signal": 0.0,
            }
        )

    result = evaluate_accuracy_target(
        pd.DataFrame(rows),
        candidate_config={"weights": {"new_signal": 0.05, "old_signal": 0.01}, "hold_bias": 0.0},
        production_config={"weights": {"old_signal": 0.01}, "hold_bias": 0.0},
        optimize_ratio=0.5,
        thresholds=TargetThresholds(min_accuracy_delta_pp=1.0),
    )

    assert result["passed"] is True
    assert result["checks"]["accuracy_delta_at_least_target"] is True
    assert result["checks"]["buy_precision_not_down"] is True
    assert result["checks"]["sell_precision_not_down"] is True
    assert result["checks"]["all_candidate_factor_weights_nonzero"] is True
    assert result["price_gate"]["passed"] is True
    assert result["proposed_golden_config"]["weights"]["new_signal"] == 0.05


def test_accuracy_target_gate_blocks_zero_weight_candidate():
    rows = [
        {
            "stock": "2330",
            "date": f"2026-06-{idx + 1:02d}",
            "y_true": 1,
            "p_buy": 0.44,
            "p_hold": 0.48,
            "p_sell": 0.08,
            "base_pred": 0,
            "factor__new_signal": 1.0,
        }
        for idx in range(8)
    ]

    result = evaluate_accuracy_target(
        pd.DataFrame(rows),
        candidate_config={"weights": {"new_signal": 0.0}, "hold_bias": 0.0},
        production_config={"weights": {"new_signal": 0.01}, "hold_bias": 0.0},
        optimize_ratio=0.5,
        thresholds=TargetThresholds(min_accuracy_delta_pp=1.0),
    )

    assert result["passed"] is False
    assert result["checks"]["all_candidate_factor_weights_nonzero"] is False
    assert result["proposed_golden_config"] is None