from __future__ import annotations

from statistics import mean, median
from typing import Any

WEIGHTS = {
    "breadth": 20,
    "limit_ecology": 25,
    "profit_effect": 30,
    "ladder_structure": 15,
    "liquidity": 10,
}


def calculate_sentiment(snapshot: dict[str, Any], history: list[dict[str, Any]]) -> dict[str, Any]:
    stats = _stats(snapshot)
    historical = [_stats(item) for item in history[-250:]]
    breadth = _clamp(stats["breadth_ratio"])
    limit_strength = _adaptive(
        stats["limit_up"], _linear(stats["limit_up"], 10, 100), _series(historical, "limit_up")
    )
    down_pressure = _adaptive(
        stats["limit_down"], _linear(stats["limit_down"], 0, 50), _series(historical, "limit_down")
    )
    down_relief = 100 - down_pressure
    seal_quality = _linear(stats["seal_rate"], 35, 90)
    ecology = limit_strength * 0.35 + seal_quality * 0.35 + down_relief * 0.30
    systemic_health = breadth * 0.60 + down_relief * 0.40
    gate = 1 if systemic_health >= 35 else 0.35 + systemic_health / 35 * 0.65

    profit = _profit(stats)
    max_height = _adaptive(
        stats["max_height"],
        _linear(stats["max_height"], 1, 7),
        _series(historical, "max_height"),
    )
    continuation = (stats["second_board"] + stats["three_plus"]) / max(stats["limit_up"], 1) * 100
    three_density = stats["three_plus"] / max(stats["limit_up"], 1) * 100
    three_score = _adaptive(
        stats["three_plus"], _clamp(three_density * 5), _series(historical, "three_plus")
    )
    ladder = (
        max_height * 0.30
        + _clamp(continuation * 3) * 0.25
        + three_score * 0.25
        + stats["ladder_completeness"] * 0.20
    )

    prior_amounts = [item["amount"] for item in historical[-20:] if item["amount"] > 0]
    baseline = mean(prior_amounts) if prior_amounts else stats["amount"] or 1
    amount_ratio = stats["amount"] / max(baseline, 1)
    amount_score = _clamp(50 + (amount_ratio - 1) * 100)
    limit_share = stats["limit_amount"] / max(stats["amount"], 1) * 100
    liquidity = amount_score * 0.70 + _clamp(limit_share * 20) * 0.30

    components = {
        "breadth": breadth,
        "limit_ecology": ecology,
        "profit_effect": profit,
        "ladder_structure": ladder,
        "liquidity": liquidity,
    }
    score = round(sum(components[key] * weight / 100 for key, weight in WEIGHTS.items()) * gate)
    extreme = stats["breadth_ratio"] <= 15 and stats["limit_down"] >= 100
    if extreme:
        score = min(score, 15)
    elif stats["breadth_ratio"] <= 25 and stats["limit_down"] >= 50:
        score = min(score, 24)

    prior_sentiments = [item.get("sentiment") or {} for item in history[-3:]]
    prior_scores = [
        float(item["score"]) for item in prior_sentiments if item.get("score") is not None
    ]
    momentum = score - mean(prior_scores) if prior_scores else 0
    direction = "升温" if momentum > 3 else "降温" if momentum < -3 else "持平"
    previous = prior_sentiments[-1] if prior_sentiments else None
    day_change = (
        score - float(previous["score"]) if previous and previous.get("score") is not None else 0
    )
    signal = _phase_signal(score, momentum, profit)
    phase, reason = _phase(
        previous, score, day_change, systemic_health, profit, ecology, signal, extreme
    )
    fermentation_ready = signal == "发酵" and score >= 45 and profit >= 45 and systemic_health >= 35
    previous_count = int(previous.get("fermentation_signal_count") or 0) if previous else 0
    fermentation_count = previous_count + 1 if fermentation_ready else 0
    history_days = len(history)
    confidence = min(95, round(55 + min(history_days, 20) * 1.25 + (15 if phase == signal else 7)))
    labels = {
        "breadth": "市场宽度",
        "limit_ecology": "涨停生态",
        "profit_effect": "赚钱效应",
        "ladder_structure": "连板结构",
        "liquidity": "成交活跃度",
    }
    return {
        "score": score,
        "label": _label(score),
        "direction": direction,
        "momentum": round(momentum, 1),
        "day_change": round(day_change, 1),
        "phase": phase,
        "phase_signal": signal,
        "transition_reason": reason,
        "fermentation_signal_count": fermentation_count,
        "confidence": confidence,
        "history_days": history_days,
        "systemic_health": round(systemic_health, 1),
        "components": [
            {
                "key": key,
                "label": labels[key],
                "score": round(value, 1),
                "weight": WEIGHTS[key],
            }
            for key, value in components.items()
        ],
        "stats": stats,
    }


def _stats(snapshot: dict[str, Any]) -> dict[str, float]:
    overview = snapshot.get("overview") or {}
    limits = snapshot.get("limits") or []
    yesterday = snapshot.get("yesterday_limits") or []
    streaks = [max(1, int(_number(row.get("streak"), 1))) for row in limits]
    levels = set(streaks)
    max_height = max(streaks, default=0)
    active = _number(overview.get("up_count")) + _number(overview.get("down_count"))
    changes = [_number(row.get("current_change")) for row in yesterday]
    yesterday_count = len(yesterday)
    return {
        "breadth_ratio": _number(overview.get("up_count")) / max(active, 1) * 100,
        "limit_up": _number(overview.get("limit_up")),
        "limit_down": _number(overview.get("limit_down")),
        "broken": _number(overview.get("broken")),
        "seal_rate": _number(overview.get("seal_rate")),
        "amount": _number(overview.get("amount")),
        "limit_amount": sum(_number(row.get("amount")) for row in limits),
        "second_board": sum(streak == 2 for streak in streaks),
        "three_plus": sum(streak >= 3 for streak in streaks),
        "max_height": max_height,
        "ladder_completeness": (
            sum(level in levels for level in range(1, max_height + 1)) / max_height * 100
            if max_height
            else 0
        ),
        "yesterday_count": yesterday_count,
        "positive_rate": sum(change > 0 for change in changes) / max(yesterday_count, 1) * 100,
        "advance_rate": sum(row.get("outcome") == "晋级" for row in yesterday)
        / max(yesterday_count, 1)
        * 100,
        "average_change": mean(changes) if changes else 0,
        "median_change": median(changes) if changes else 0,
        "severe_loss_rate": sum(change <= -5 for change in changes) / max(yesterday_count, 1) * 100,
        "previous_down_rate": sum(row.get("outcome") == "跌停" for row in yesterday)
        / max(yesterday_count, 1)
        * 100,
    }


def _profit(stats: dict[str, float]) -> float:
    if not stats["yesterday_count"]:
        return 50
    median_score = _clamp(50 + stats["median_change"] * 7)
    average_score = _clamp(50 + stats["average_change"] * 6)
    advance_score = _clamp(stats["advance_rate"] * 2.5)
    loss_safety = _clamp(100 - stats["severe_loss_rate"] * 3)
    down_safety = _clamp(100 - stats["previous_down_rate"] * 7)
    tail = loss_safety * 0.70 + down_safety * 0.30
    return (
        stats["positive_rate"] * 0.30
        + median_score * 0.25
        + average_score * 0.10
        + advance_score * 0.20
        + tail * 0.15
    )


def _phase_signal(score: float, momentum: float, profit: float) -> str:
    if score < 25:
        return "修复" if momentum > 3 else "冰点"
    if score < 45:
        return "修复" if momentum > 3 else "退潮"
    if score >= 80:
        return "高潮" if momentum >= -2 and profit >= 60 else "分化"
    if score >= 65:
        return "分化" if momentum < -3 or profit < 50 else "发酵"
    if momentum < -5:
        return "退潮"
    return "发酵" if momentum >= 0 and profit >= 45 else "分化"


def _phase(previous, score, change, health, profit, ecology, signal, extreme):
    if not previous:
        return signal, "首个连续交易日，采用原始阶段信号"
    prior = str(previous.get("phase") or signal)
    if extreme:
        return "冰点", "市场宽度与跌停数量触发极端冰点"
    recovery = change >= 6 and score >= 25 and health >= 24
    climax = score >= 80 and profit >= 60 and health >= 60 and ecology >= 70
    if prior in {"冰点", "退潮"}:
        if score < 25:
            return "冰点", "市场仍处于冰点区间"
        return ("修复", "出现有效回升") if recovery else (prior, "尚未形成有效修复")
    if prior == "修复":
        if score < 25:
            return "冰点", "修复失败并跌入冰点"
        if change <= -6 and score < 45:
            return "退潮", "修复失败且显著降温"
        prior_signal = int(previous.get("fermentation_signal_count") or 0)
        if signal == "发酵" and prior_signal >= 1:
            return "发酵", "发酵条件连续两个交易日成立"
        return "修复", "修复延续，等待发酵确认"
    if prior == "发酵":
        if score < 45 and (change < 0 or health < 35):
            return "退潮", "温度与系统健康度转弱"
        if climax:
            return "高潮", "温度、赚钱效应与涨停生态达到高潮条件"
        return ("分化", "发酵阶段出现降温") if signal in {"分化", "退潮"} else ("发酵", "发酵延续")
    if prior == "高潮":
        if climax:
            return "高潮", "高潮条件继续成立"
        return ("退潮", "风险快速释放") if score < 45 or health < 30 else ("分化", "高潮条件消退")
    if score < 25:
        return "冰点", "分化继续恶化至冰点"
    if score < 45 or health < 30:
        return "退潮", "分化后继续转弱"
    return "分化", "分化延续，等待方向确认"


def _label(score: float) -> str:
    if score >= 80:
        return "情绪高涨"
    if score >= 60:
        return "情绪偏强"
    if score >= 40:
        return "情绪中性"
    if score >= 20:
        return "情绪偏弱"
    return "情绪冰点"


def _series(rows: list[dict[str, float]], key: str) -> list[float]:
    return [row[key] for row in rows]


def _adaptive(value: float, fixed: float, history: list[float]) -> float:
    if len(history) < 20:
        return fixed
    below = sum(item < value for item in history[-250:])
    equal = sum(item == value for item in history[-250:])
    percentile = (below + equal * 0.5) / len(history[-250:]) * 100
    return fixed * 0.25 + percentile * 0.75


def _linear(value: float, low: float, high: float) -> float:
    return _clamp((value - low) / (high - low) * 100) if high > low else 50


def _clamp(value: float) -> float:
    return min(100, max(0, value))


def _number(value: Any, default: float = 0.0) -> float:
    try:
        number = float(value)
        return number if number == number else default
    except (TypeError, ValueError):
        return default