//! Activity Bonus CLI Configuration Tests (Wave 16S V13)
//!
//! Validates CLI parameter handling, boundary conditions, and integration
//! with ExtrinsicRewardCalculator.

use ml::dqn::action_space::{ExposureLevel, FactoredAction, OrderType, Urgency};
use ml::dqn::reward_elite::ExtrinsicRewardCalculator;

/// Test default activity bonus values (0.10 weight, 0.05 bonus, -0.10 penalty)
#[test]
fn test_default_activity_bonus_values() {
    let calc = ExtrinsicRewardCalculator::new();

    // BUY action (zero P&L, isolate activity bonus)
    let buy_action = FactoredAction::new(ExposureLevel::Long100, OrderType::Market, Urgency::Aggressive);
    let reward_buy = calc.clone().calculate_extrinsic_reward(
        buy_action,
        100.0, // entry_price
        100.0, // exit_price (no P&L)
        10.0,  // position_size
        10000.0, // portfolio_value
        0.0,   // max_drawdown
    );

    // HOLD action (zero P&L, isolate activity penalty)
    let hold_action = FactoredAction::new(ExposureLevel::Flat, OrderType::Market, Urgency::Normal);
    let reward_hold = calc.clone().calculate_extrinsic_reward(
        hold_action,
        100.0,
        100.0,
        10.0,
        10000.0,
        0.0,
    );

    // Expected: BUY = 0.10 * 0.05 = 0.005
    // Expected: HOLD = 0.10 * (-0.10) = -0.01
    assert!((reward_buy - 0.005).abs() < 1e-6, "BUY reward should be 0.005, got {}", reward_buy);
    assert!((reward_hold - (-0.01)).abs() < 1e-6, "HOLD reward should be -0.01, got {}", reward_hold);

    // Difference should be 0.015 (0.10 * (0.05 - (-0.10)))
    let diff = reward_buy - reward_hold;
    assert!((diff - 0.015).abs() < 1e-6, "Difference should be 0.015, got {}", diff);
}

/// Test custom activity bonus values
#[test]
fn test_custom_activity_bonus_values() {
    let calc = ExtrinsicRewardCalculator::with_config(
        100,  // sharpe_window
        0.20, // activity_bonus_weight (20% instead of 10%)
        0.10, // activity_bonus_value (0.10 instead of 0.05)
        -0.20, // activity_penalty_value (-0.20 instead of -0.10)
    );

    let buy_action = FactoredAction::new(ExposureLevel::Long100, OrderType::Market, Urgency::Aggressive);
    let hold_action = FactoredAction::new(ExposureLevel::Flat, OrderType::Market, Urgency::Normal);

    let reward_buy = calc.clone().calculate_extrinsic_reward(
        buy_action,
        100.0,
        100.0,
        10.0,
        10000.0,
        0.0,
    );

    let reward_hold = calc.clone().calculate_extrinsic_reward(
        hold_action,
        100.0,
        100.0,
        10.0,
        10000.0,
        0.0,
    );

    // Expected: BUY = 0.20 * 0.10 = 0.02
    // Expected: HOLD = 0.20 * (-0.20) = -0.04
    assert!((reward_buy - 0.02).abs() < 1e-6, "BUY reward should be 0.02, got {}", reward_buy);
    assert!((reward_hold - (-0.04)).abs() < 1e-6, "HOLD reward should be -0.04, got {}", reward_hold);
}

/// Test disabled activity bonus (weight = 0.0)
#[test]
fn test_disabled_activity_bonus() {
    let calc = ExtrinsicRewardCalculator::with_config(
        100,  // sharpe_window
        0.0,  // activity_bonus_weight (disabled)
        0.05, // activity_bonus_value (irrelevant when disabled)
        -0.10, // activity_penalty_value (irrelevant when disabled)
    );

    let buy_action = FactoredAction::new(ExposureLevel::Long100, OrderType::Market, Urgency::Aggressive);
    let hold_action = FactoredAction::new(ExposureLevel::Flat, OrderType::Market, Urgency::Normal);

    let reward_buy = calc.clone().calculate_extrinsic_reward(
        buy_action,
        100.0,
        100.0,
        10.0,
        10000.0,
        0.0,
    );

    let reward_hold = calc.clone().calculate_extrinsic_reward(
        hold_action,
        100.0,
        100.0,
        10.0,
        10000.0,
        0.0,
    );

    // Expected: both should be 0.0 (no P&L, no activity bonus)
    assert!(reward_buy.abs() < 1e-6, "BUY reward should be 0.0 when activity bonus disabled, got {}", reward_buy);
    assert!(reward_hold.abs() < 1e-6, "HOLD reward should be 0.0 when activity bonus disabled, got {}", reward_hold);
}

/// Test boundary values: activity_bonus_weight = 1.0 (maximum)
#[test]
fn test_max_activity_bonus_weight() {
    let calc = ExtrinsicRewardCalculator::with_config(
        100,  // sharpe_window
        1.0,  // activity_bonus_weight (100%)
        0.05,
        -0.10,
    );

    let buy_action = FactoredAction::new(ExposureLevel::Long100, OrderType::Market, Urgency::Aggressive);
    let reward_buy = calc.clone().calculate_extrinsic_reward(
        buy_action,
        100.0,
        100.0,
        10.0,
        10000.0,
        0.0,
    );

    // Expected: BUY = 1.0 * 0.05 = 0.05 (activity bonus dominates)
    assert!((reward_buy - 0.05).abs() < 1e-6, "BUY reward should be 0.05 with 100% weight, got {}", reward_buy);
}

/// Test boundary values: negative bonus (penalize BUY/SELL, reward HOLD)
#[test]
fn test_inverted_activity_bonus() {
    let calc = ExtrinsicRewardCalculator::with_config(
        100,   // sharpe_window
        0.10,  // activity_bonus_weight
        -0.05, // activity_bonus_value (negative for BUY/SELL)
        0.10,  // activity_penalty_value (positive for HOLD)
    );

    let buy_action = FactoredAction::new(ExposureLevel::Long100, OrderType::Market, Urgency::Aggressive);
    let hold_action = FactoredAction::new(ExposureLevel::Flat, OrderType::Market, Urgency::Normal);

    let reward_buy = calc.clone().calculate_extrinsic_reward(
        buy_action,
        100.0,
        100.0,
        10.0,
        10000.0,
        0.0,
    );

    let reward_hold = calc.clone().calculate_extrinsic_reward(
        hold_action,
        100.0,
        100.0,
        10.0,
        10000.0,
        0.0,
    );

    // Expected: BUY = 0.10 * (-0.05) = -0.005 (penalized)
    // Expected: HOLD = 0.10 * 0.10 = 0.01 (rewarded)
    assert!((reward_buy - (-0.005)).abs() < 1e-6, "BUY reward should be -0.005, got {}", reward_buy);
    assert!((reward_hold - 0.01).abs() < 1e-6, "HOLD reward should be 0.01, got {}", reward_hold);
}

/// Test all exposure levels get BUY/SELL bonus (except Flat)
#[test]
fn test_all_exposure_levels() {
    let calc = ExtrinsicRewardCalculator::new();

    let exposure_levels = vec![
        (ExposureLevel::Short100, "Short100"),
        (ExposureLevel::Short50, "Short50"),
        (ExposureLevel::Flat, "Flat"),
        (ExposureLevel::Long50, "Long50"),
        (ExposureLevel::Long100, "Long100"),
    ];

    for (exposure, name) in exposure_levels {
        let action = FactoredAction::new(exposure, OrderType::Market, Urgency::Normal);
        let reward = calc.clone().calculate_extrinsic_reward(
            action,
            100.0,
            100.0,
            10.0,
            10000.0,
            0.0,
        );

        if matches!(exposure, ExposureLevel::Flat) {
            // Flat should get HOLD penalty
            assert!(
                (reward - (-0.01)).abs() < 1e-6,
                "{} should get HOLD penalty (-0.01), got {}",
                name,
                reward
            );
        } else {
            // All others should get BUY/SELL bonus
            assert!(
                (reward - 0.005).abs() < 1e-6,
                "{} should get BUY/SELL bonus (0.005), got {}",
                name,
                reward
            );
        }
    }
}

/// Test activity bonus with non-zero P&L
#[test]
fn test_activity_bonus_with_pnl() {
    let calc = ExtrinsicRewardCalculator::new();

    // BUY with 5% profit
    let buy_action = FactoredAction::new(ExposureLevel::Long100, OrderType::Market, Urgency::Aggressive);
    let reward = calc.clone().calculate_extrinsic_reward(
        buy_action,
        100.0,  // entry
        105.0,  // exit (+5% profit)
        10.0,   // position_size
        10000.0, // portfolio_value
        0.0,
    );

    // Expected calculation:
    // P&L = (105 - 100) * 10 = 50.0
    // Normalized P&L = 50.0 / 10000.0 = 0.005
    // P&L component = 0.40 * 0.005 = 0.002
    // Sharpe = 0.0 (first call, insufficient data)
    // Drawdown = 0.0
    // Activity bonus = 0.10 * 0.05 = 0.005
    // Total = 0.002 + 0.0 + 0.0 + 0.005 = 0.007
    assert!((reward - 0.007).abs() < 1e-6, "Expected 0.007, got {}", reward);
}

/// Test activity bonus does not interfere with other reward components
#[test]
fn test_activity_bonus_independence() {
    let calc_default = ExtrinsicRewardCalculator::new(); // 10% activity bonus
    let calc_disabled = ExtrinsicRewardCalculator::with_config(100, 0.0, 0.0, 0.0); // 0% activity bonus

    let buy_action = FactoredAction::new(ExposureLevel::Long100, OrderType::Market, Urgency::Aggressive);

    // Same profitable trade
    let reward_default = calc_default.clone().calculate_extrinsic_reward(
        buy_action,
        100.0,
        110.0, // 10% profit
        100.0,
        10000.0,
        0.0,
    );

    let reward_disabled = calc_disabled.clone().calculate_extrinsic_reward(
        buy_action,
        100.0,
        110.0,
        100.0,
        10000.0,
        0.0,
    );

    // Difference should be exactly the activity bonus contribution
    // P&L = 1000.0, normalized = 0.10
    // P&L component = 0.40 * 0.10 = 0.04
    // Activity bonus = 0.10 * 0.05 = 0.005
    // Default total = 0.04 + 0.005 = 0.045
    // Disabled total = 0.04
    let diff = reward_default - reward_disabled;
    assert!((diff - 0.005).abs() < 1e-6, "Difference should be 0.005 (activity bonus), got {}", diff);
}

/// Test validation: activity_bonus_weight out of range (should be validated at CLI layer)
#[test]
#[should_panic(expected = "Sharpe window must be >= 2")]
fn test_invalid_sharpe_window() {
    let _calc = ExtrinsicRewardCalculator::with_config(
        1,    // Invalid: sharpe_window < 2
        0.10,
        0.05,
        -0.10,
    );
}

/// Test backward compatibility: existing code using `new()` gets default values
#[test]
fn test_backward_compatibility() {
    let calc_new = ExtrinsicRewardCalculator::new();
    let calc_explicit = ExtrinsicRewardCalculator::with_config(100, 0.10, 0.05, -0.10);

    let buy_action = FactoredAction::new(ExposureLevel::Long100, OrderType::Market, Urgency::Aggressive);

    let reward_new = calc_new.clone().calculate_extrinsic_reward(
        buy_action,
        100.0,
        100.0,
        10.0,
        10000.0,
        0.0,
    );

    let reward_explicit = calc_explicit.clone().calculate_extrinsic_reward(
        buy_action,
        100.0,
        100.0,
        10.0,
        10000.0,
        0.0,
    );

    // Should be identical
    assert_eq!(reward_new, reward_explicit, "new() and explicit defaults should produce identical rewards");
}