"""Property-based tests for the TradingEngine core decision loop.

Feature: autonomous-trading-engine

Tests properties 27, 28, 16, and 18 from the design specification,
covering recommendation deduplication, decision record completeness,
multiple declining positions halt, and maximum open positions enforcement.
"""
from __future__ import annotations

from datetime import datetime, timezone
from zoneinfo import ZoneInfo

from hypothesis import given, settings, assume
from hypothesis import strategies as st

from services.shared.config import TradingConfig
from services.trading.correlation import CorrelationMatrix
from services.trading.engine import TradingEngine
from services.trading.models import (
    CircuitBreakerState,
    OpenPosition,
    PortfolioState,
    RiskTierConfig,
)


# ---------------------------------------------------------------------------
# Helpers
# ---------------------------------------------------------------------------

ET = ZoneInfo("America/New_York")

# A valid trading window datetime: Wednesday 10:00 AM ET
VALID_TRADING_DT = datetime(2025, 1, 8, 10, 0, 0, tzinfo=ET)


def _make_engine() -> TradingEngine:
    """Create a TradingEngine with default TradingConfig and no pool/redis."""
    return TradingEngine(pool=None, redis=None, config=TradingConfig())


def _moderate_risk_tier() -> RiskTierConfig:
    """Return a moderate risk tier with reasonable defaults for testing."""
    return RiskTierConfig(
        name="moderate",
        min_confidence=0.55,
        max_position_pct=0.10,
        stop_loss_atr_multiplier=2.0,
        reward_risk_ratio=1.5,
        max_sector_pct=0.30,
        max_portfolio_heat=0.20,
    )


def _inactive_cb() -> CircuitBreakerState:
    """Return an inactive circuit breaker state."""
    return CircuitBreakerState(active=False)


def _empty_correlation_matrix() -> CorrelationMatrix:
    """Return an empty correlation matrix."""
    return CorrelationMatrix()


def _base_portfolio(
    active_pool: float = 5000.0,
    positions: list | None = None,
    portfolio_heat: float = 0.0,
    open_position_count: int = 0,
) -> PortfolioState:
    """Return a PortfolioState with sensible defaults for testing."""
    return PortfolioState(
        positions=positions or [],
        total_value=active_pool + 500.0,
        cash=active_pool,
        active_pool=active_pool,
        reserve_pool=500.0,
        sector_exposure={},
        portfolio_heat=portfolio_heat,
        open_position_count=open_position_count,
    )


def _make_recommendation(
    rec_id: str = "rec-001",
    ticker: str = "AAPL",
    confidence: float = 0.80,
    sector: str = "Technology",
    current_price: float = 10.0,
    action: str = "buy",
) -> dict:
    """Build a recommendation dict suitable for evaluate_recommendation."""
    return {
        "recommendation_id": rec_id,
        "ticker": ticker,
        "confidence": confidence,
        "sector": sector,
        "current_price": current_price,
        "action": action,
    }


# ---------------------------------------------------------------------------
# Hypothesis strategies
# ---------------------------------------------------------------------------

def _recommendation_id_strategy() -> st.SearchStrategy[str]:
    """Generate random recommendation IDs."""
    return st.text(
        alphabet=st.sampled_from("abcdefghijklmnopqrstuvwxyz0123456789-"),
        min_size=5,
        max_size=20,
    ).filter(lambda s: len(s.strip()) > 0)


def _open_position_strategy(
    unrealized_pnl: st.SearchStrategy[float] | None = None,
) -> st.SearchStrategy[OpenPosition]:
    """Generate random OpenPosition objects."""
    return st.builds(
        OpenPosition,
        ticker=st.from_regex(r"[A-Z]{3,5}", fullmatch=True),
        quantity=st.integers(min_value=1, max_value=100),
        entry_price=st.floats(min_value=5.0, max_value=500.0, allow_nan=False, allow_infinity=False),
        current_price=st.floats(min_value=5.0, max_value=500.0, allow_nan=False, allow_infinity=False),
        unrealized_pnl=unrealized_pnl if unrealized_pnl is not None else st.floats(
            min_value=-1000.0, max_value=1000.0, allow_nan=False, allow_infinity=False
        ),
        market_value=st.floats(min_value=10.0, max_value=5000.0, allow_nan=False, allow_infinity=False),
        sector=st.sampled_from(["Technology", "Healthcare", "Energy", "Financials", "Consumer"]),
        stop_loss_price=st.floats(min_value=1.0, max_value=400.0, allow_nan=False, allow_infinity=False),
        take_profit_price=st.floats(min_value=10.0, max_value=1000.0, allow_nan=False, allow_infinity=False),
        signal_confidence=st.floats(min_value=0.0, max_value=1.0, allow_nan=False, allow_infinity=False),
        is_micro_trade=st.just(False),
    )


# ---------------------------------------------------------------------------
# Property 27: Recommendation deduplication (idempotence)
# **Validates: Requirements 1.5**
# ---------------------------------------------------------------------------


class TestProperty27RecommendationDeduplication:
    """Property 27: Recommendation deduplication (idempotence).

    **Validates: Requirements 1.5**
    """

    @settings(max_examples=100)
    @given(rec_id=_recommendation_id_strategy())
    def test_duplicate_recommendation_produces_skip(self, rec_id: str) -> None:
        """Processing the same recommendation twice produces a skip on the second call."""
        engine = _make_engine()
        risk_tier = _moderate_risk_tier()
        portfolio = _base_portfolio(active_pool=5000.0)
        cb_state = _inactive_cb()
        corr_matrix = _empty_correlation_matrix()

        rec = _make_recommendation(
            rec_id=rec_id,
            ticker="AAPL",
            confidence=0.80,
            sector="Technology",
            current_price=10.0,
        )

        # First processing — should produce an "act" decision
        decision1 = engine.evaluate_recommendation(
            rec=rec,
            portfolio_state=portfolio,
            risk_tier=risk_tier,
            circuit_breaker_state=cb_state,
            correlation_matrix=corr_matrix,
            earnings_calendar={},
            now=VALID_TRADING_DT,
        )
        assert decision1.decision == "act", (
            f"First processing should produce 'act', got '{decision1.decision}' "
            f"with skip_reason={decision1.skip_reason}"
        )

        # Second processing — should produce a "skip" with duplicate reason
        decision2 = engine.evaluate_recommendation(
            rec=rec,
            portfolio_state=portfolio,
            risk_tier=risk_tier,
            circuit_breaker_state=cb_state,
            correlation_matrix=corr_matrix,
            earnings_calendar={},
            now=VALID_TRADING_DT,
        )
        assert decision2.decision == "skip"
        assert decision2.skip_reason == "duplicate_recommendation"

    @settings(max_examples=100)
    @given(
        rec_id_a=_recommendation_id_strategy(),
        rec_id_b=_recommendation_id_strategy(),
    )
    def test_different_recommendations_not_deduplicated(
        self, rec_id_a: str, rec_id_b: str,
    ) -> None:
        """Different recommendation IDs are processed independently."""
        assume(rec_id_a != rec_id_b)

        engine = _make_engine()
        risk_tier = _moderate_risk_tier()
        portfolio = _base_portfolio(active_pool=5000.0)
        cb_state = _inactive_cb()
        corr_matrix = _empty_correlation_matrix()

        rec_a = _make_recommendation(rec_id=rec_id_a, ticker="AAPL", current_price=10.0)
        rec_b = _make_recommendation(rec_id=rec_id_b, ticker="MSFT", current_price=10.0)

        decision_a = engine.evaluate_recommendation(
            rec=rec_a,
            portfolio_state=portfolio,
            risk_tier=risk_tier,
            circuit_breaker_state=cb_state,
            correlation_matrix=corr_matrix,
            earnings_calendar={},
            now=VALID_TRADING_DT,
        )

        decision_b = engine.evaluate_recommendation(
            rec=rec_b,
            portfolio_state=portfolio,
            risk_tier=risk_tier,
            circuit_breaker_state=cb_state,
            correlation_matrix=corr_matrix,
            earnings_calendar={},
            now=VALID_TRADING_DT,
        )

        # Both should be "act" (not deduplicated against each other)
        assert decision_a.decision == "act"
        assert decision_b.decision == "act"


# ---------------------------------------------------------------------------
# Property 28: Trading decision record completeness and traceability
# **Validates: Requirements 1.4, 17.1, 17.2**
# ---------------------------------------------------------------------------


class TestProperty28DecisionRecordCompleteness:
    """Property 28: Trading decision record completeness and traceability.

    **Validates: Requirements 1.4, 17.1, 17.2**
    """

    @settings(max_examples=100)
    @given(
        confidence=st.floats(min_value=0.60, max_value=0.99, allow_nan=False, allow_infinity=False),
        ticker=st.from_regex(r"[A-Z]{3,5}", fullmatch=True),
        price=st.floats(min_value=1.0, max_value=50.0, allow_nan=False, allow_infinity=False),
    )
    def test_act_decision_has_all_required_fields(
        self, confidence: float, ticker: str, price: float,
    ) -> None:
        """Act decisions have all required fields including position size."""
        engine = _make_engine()
        risk_tier = _moderate_risk_tier()
        portfolio = _base_portfolio(active_pool=5000.0)
        cb_state = _inactive_cb()
        corr_matrix = _empty_correlation_matrix()

        rec = _make_recommendation(
            rec_id=f"rec-{ticker}",
            ticker=ticker,
            confidence=confidence,
            current_price=price,
        )

        decision = engine.evaluate_recommendation(
            rec=rec,
            portfolio_state=portfolio,
            risk_tier=risk_tier,
            circuit_breaker_state=cb_state,
            correlation_matrix=corr_matrix,
            earnings_calendar={},
            now=VALID_TRADING_DT,
        )

        # All required fields must be present and non-None
        assert decision.id is not None and len(decision.id) > 0
        assert decision.recommendation_id == f"rec-{ticker}"
        assert decision.decision in ("act", "skip")
        assert decision.ticker == ticker
        assert decision.risk_tier_at_decision == "moderate"
        assert decision.circuit_breaker_status in ("active", "inactive")
        assert decision.decision_trace is not None
        assert isinstance(decision.decision_trace, dict)
        assert decision.created_at is not None

        if decision.decision == "act":
            assert decision.computed_position_size is not None
            assert decision.computed_position_size > 0
            assert decision.computed_share_quantity is not None
            assert decision.computed_share_quantity > 0
            assert decision.skip_reason is None

    @settings(max_examples=100)
    @given(
        confidence=st.floats(min_value=0.01, max_value=0.50, allow_nan=False, allow_infinity=False),
        ticker=st.from_regex(r"[A-Z]{3,5}", fullmatch=True),
    )
    def test_skip_decision_has_skip_reason(
        self, confidence: float, ticker: str,
    ) -> None:
        """Skip decisions have skip_reason set."""
        engine = _make_engine()
        risk_tier = _moderate_risk_tier()
        portfolio = _base_portfolio(active_pool=5000.0)
        cb_state = _inactive_cb()
        corr_matrix = _empty_correlation_matrix()

        # Use low confidence to trigger a skip
        assume(confidence < risk_tier.min_confidence)

        rec = _make_recommendation(
            rec_id=f"rec-{ticker}",
            ticker=ticker,
            confidence=confidence,
            current_price=10.0,
        )

        decision = engine.evaluate_recommendation(
            rec=rec,
            portfolio_state=portfolio,
            risk_tier=risk_tier,
            circuit_breaker_state=cb_state,
            correlation_matrix=corr_matrix,
            earnings_calendar={},
            now=VALID_TRADING_DT,
        )

        assert decision.decision == "skip"
        assert decision.skip_reason is not None
        assert len(decision.skip_reason) > 0

        # Required fields still present on skip decisions
        assert decision.id is not None
        assert decision.recommendation_id is not None
        assert decision.ticker == ticker
        assert decision.risk_tier_at_decision == "moderate"
        assert decision.circuit_breaker_status == "inactive"
        assert decision.decision_trace is not None
        assert decision.created_at is not None

    @settings(max_examples=100)
    @given(
        confidence=st.floats(min_value=0.60, max_value=0.99, allow_nan=False, allow_infinity=False),
        ticker=st.from_regex(r"[A-Z]{3,5}", fullmatch=True),
        price=st.floats(min_value=1.0, max_value=50.0, allow_nan=False, allow_infinity=False),
    )
    def test_decision_trace_contains_reasoning(
        self, confidence: float, ticker: str, price: float,
    ) -> None:
        """Decision trace contains a reasoning list for traceability."""
        engine = _make_engine()
        risk_tier = _moderate_risk_tier()
        portfolio = _base_portfolio(active_pool=5000.0)
        cb_state = _inactive_cb()
        corr_matrix = _empty_correlation_matrix()

        rec = _make_recommendation(
            rec_id=f"rec-trace-{ticker}",
            ticker=ticker,
            confidence=confidence,
            current_price=price,
        )

        decision = engine.evaluate_recommendation(
            rec=rec,
            portfolio_state=portfolio,
            risk_tier=risk_tier,
            circuit_breaker_state=cb_state,
            correlation_matrix=corr_matrix,
            earnings_calendar={},
            now=VALID_TRADING_DT,
        )

        assert "reasoning" in decision.decision_trace
        assert isinstance(decision.decision_trace["reasoning"], list)
        assert len(decision.decision_trace["reasoning"]) > 0


# ---------------------------------------------------------------------------
# Property 16: Multiple declining positions halts new entries
# **Validates: Requirements 7.5**
# ---------------------------------------------------------------------------


class TestProperty16DecliningPositionsHalt:
    """Property 16: Multiple declining positions halts new entries.

    **Validates: Requirements 7.5**
    """

    @settings(max_examples=100)
    @given(
        total_positions=st.integers(min_value=4, max_value=20),
        declining_fraction=st.floats(min_value=0.55, max_value=1.0, allow_nan=False, allow_infinity=False),
    )
    def test_entries_halted_when_majority_declining(
        self, total_positions: int, declining_fraction: float,
    ) -> None:
        """New entries halted when > 50% of positions have > 2% negative unrealized P&L."""
        declining_count = int(total_positions * declining_fraction)
        assume(declining_count > total_positions * 0.50)  # strictly > 50%
        non_declining_count = total_positions - declining_count

        positions: list[OpenPosition] = []

        # Declining positions: unrealized_pnl is negative and > 2% of entry value
        for i in range(declining_count):
            entry_price = 100.0
            quantity = 10
            entry_value = entry_price * quantity
            # Loss > 2% of entry value
            loss = entry_value * 0.03  # 3% loss
            positions.append(OpenPosition(
                ticker=f"DEC{i}",
                quantity=quantity,
                entry_price=entry_price,
                current_price=entry_price - (loss / quantity),
                unrealized_pnl=-loss,
                market_value=entry_value - loss,
                sector="Technology",
                stop_loss_price=90.0,
                take_profit_price=120.0,
                signal_confidence=0.6,
            ))

        # Non-declining positions: small positive or flat P&L
        for i in range(non_declining_count):
            entry_price = 100.0
            quantity = 10
            positions.append(OpenPosition(
                ticker=f"OK{i}",
                quantity=quantity,
                entry_price=entry_price,
                current_price=entry_price + 1.0,
                unrealized_pnl=10.0,
                market_value=1010.0,
                sector="Healthcare",
                stop_loss_price=90.0,
                take_profit_price=120.0,
                signal_confidence=0.7,
            ))

        engine = _make_engine()

        # Verify the check_declining_positions method returns True
        assert engine.check_declining_positions(positions) is True

        # Verify the full engine skips new entries
        risk_tier = _moderate_risk_tier()
        portfolio = _base_portfolio(
            active_pool=5000.0,
            positions=positions,
            open_position_count=total_positions,
        )
        cb_state = _inactive_cb()
        corr_matrix = _empty_correlation_matrix()

        rec = _make_recommendation(
            rec_id="rec-decline-test",
            ticker="NEWSTOCK",
            confidence=0.80,
            current_price=10.0,
        )

        decision = engine.evaluate_recommendation(
            rec=rec,
            portfolio_state=portfolio,
            risk_tier=risk_tier,
            circuit_breaker_state=cb_state,
            correlation_matrix=corr_matrix,
            earnings_calendar={},
            now=VALID_TRADING_DT,
        )

        assert decision.decision == "skip"
        assert decision.skip_reason == "multiple_declining_positions"

    @settings(max_examples=100)
    @given(
        total_positions=st.integers(min_value=4, max_value=20),
        declining_fraction=st.floats(min_value=0.0, max_value=0.45, allow_nan=False, allow_infinity=False),
    )
    def test_entries_allowed_when_minority_declining(
        self, total_positions: int, declining_fraction: float,
    ) -> None:
        """Entries allowed when <= 50% of positions are declining."""
        declining_count = int(total_positions * declining_fraction)
        assume(declining_count <= total_positions * 0.50)  # at or below 50%
        non_declining_count = total_positions - declining_count

        positions: list[OpenPosition] = []

        # Declining positions
        for i in range(declining_count):
            entry_price = 100.0
            quantity = 10
            entry_value = entry_price * quantity
            loss = entry_value * 0.03
            positions.append(OpenPosition(
                ticker=f"DEC{i}",
                quantity=quantity,
                entry_price=entry_price,
                current_price=entry_price - (loss / quantity),
                unrealized_pnl=-loss,
                market_value=entry_value - loss,
                sector="Technology",
                stop_loss_price=90.0,
                take_profit_price=120.0,
                signal_confidence=0.6,
            ))

        # Non-declining positions
        for i in range(non_declining_count):
            entry_price = 100.0
            quantity = 10
            positions.append(OpenPosition(
                ticker=f"OK{i}",
                quantity=quantity,
                entry_price=entry_price,
                current_price=entry_price + 1.0,
                unrealized_pnl=10.0,
                market_value=1010.0,
                sector="Healthcare",
                stop_loss_price=90.0,
                take_profit_price=120.0,
                signal_confidence=0.7,
            ))

        engine = _make_engine()

        # Verify the check returns False (entries allowed)
        assert engine.check_declining_positions(positions) is False

    @settings(max_examples=100)
    @given(data=st.data())
    def test_empty_positions_allows_entries(self, data: st.DataObject) -> None:
        """Empty position list always allows new entries."""
        engine = _make_engine()
        assert engine.check_declining_positions([]) is False


# ---------------------------------------------------------------------------
# Property 18: Maximum open positions enforcement
# **Validates: Requirements 8.4**
# ---------------------------------------------------------------------------


class TestProperty18MaxOpenPositions:
    """Property 18: Maximum open positions enforcement.

    **Validates: Requirements 8.4**
    """

    @settings(max_examples=100)
    @given(
        max_positions=st.integers(min_value=1, max_value=20),
    )
    def test_entries_rejected_at_max_positions(self, max_positions: int) -> None:
        """New entries rejected when open_position_count >= max_open_positions."""
        engine = _make_engine()

        # Verify the check_max_positions method returns True at the limit
        assert engine.check_max_positions(max_positions, max_positions) is True

        # Also verify above the limit
        assert engine.check_max_positions(max_positions + 1, max_positions) is True

    @settings(max_examples=100)
    @given(
        max_positions=st.integers(min_value=2, max_value=20),
        open_count_offset=st.integers(min_value=1, max_value=10),
    )
    def test_entries_allowed_below_max_positions(
        self, max_positions: int, open_count_offset: int,
    ) -> None:
        """Entries allowed when open_position_count < max_open_positions."""
        open_count = max(0, max_positions - open_count_offset)
        assume(open_count < max_positions)

        engine = _make_engine()
        assert engine.check_max_positions(open_count, max_positions) is False

    @settings(max_examples=100)
    @given(
        max_positions=st.integers(min_value=1, max_value=15),
    )
    def test_full_engine_rejects_at_max_positions(self, max_positions: int) -> None:
        """Full engine evaluation rejects new entries at max positions."""
        engine = _make_engine()
        # Override the config's max_open_positions via hasattr fallback
        # The engine uses: self.config.max_open_positions if hasattr(...) else 10
        # TradingConfig doesn't have max_open_positions, so it defaults to 10.
        # We test with the default (10) by setting open_position_count = 10.
        risk_tier = _moderate_risk_tier()
        portfolio = _base_portfolio(
            active_pool=5000.0,
            open_position_count=10,  # default max is 10
        )
        cb_state = _inactive_cb()
        corr_matrix = _empty_correlation_matrix()

        rec = _make_recommendation(
            rec_id=f"rec-max-{max_positions}",
            ticker="MAXTEST",
            confidence=0.80,
            current_price=10.0,
        )

        decision = engine.evaluate_recommendation(
            rec=rec,
            portfolio_state=portfolio,
            risk_tier=risk_tier,
            circuit_breaker_state=cb_state,
            correlation_matrix=corr_matrix,
            earnings_calendar={},
            now=VALID_TRADING_DT,
        )

        assert decision.decision == "skip"
        assert decision.skip_reason == "max_positions_reached"

    @settings(max_examples=100)
    @given(
        open_count=st.integers(min_value=0, max_value=8),
    )
    def test_full_engine_allows_below_max_positions(self, open_count: int) -> None:
        """Full engine evaluation allows entries below max positions (default 10)."""
        assume(open_count < 10)  # default max is 10

        engine = _make_engine()
        risk_tier = _moderate_risk_tier()
        portfolio = _base_portfolio(
            active_pool=5000.0,
            open_position_count=open_count,
        )
        cb_state = _inactive_cb()
        corr_matrix = _empty_correlation_matrix()

        rec = _make_recommendation(
            rec_id=f"rec-below-{open_count}",
            ticker="BELOWMAX",
            confidence=0.80,
            current_price=10.0,
        )

        decision = engine.evaluate_recommendation(
            rec=rec,
            portfolio_state=portfolio,
            risk_tier=risk_tier,
            circuit_breaker_state=cb_state,
            correlation_matrix=corr_matrix,
            earnings_calendar={},
            now=VALID_TRADING_DT,
        )

        # Should not be rejected for max positions
        if decision.decision == "skip":
            assert decision.skip_reason != "max_positions_reached"