"""
LLM 增强分析服务

提供两个核心能力：
1. generate_design_report — 将分析 JSON 转换为结构化评审报告
2. recommend_parting_direction — 基于几何 + 制造约束推荐最优分型方向

适配层：OpenAI 兼容 API（支持 OpenAI / DeepSeek / vLLM / Ollama 等）
未配置 LLM 时静默降级，不影响主流程。
"""

import json
import re
from typing import Optional, Dict, Any, List

import httpx

from config.settings import settings
from utils.logger import get_logger

logger = get_logger(__name__)


_DESIGN_REPORT_SYSTEM = """你是一位资深注塑模具设计工程师，拥有 20 年模具 DFM 评审经验。
请根据提供的模具分析数据，生成一份专业的模具设计评审报告。

要求：
1. 使用中文
2. 按 "关键问题 → 工艺参数建议 → 改进建议" 结构组织
3. 技术术语准确（如：锁模力、投影面积、分型面、滑块、斜顶、拔模角、缩痕、熔接痕）
4. 每个问题标注优先级（high / medium / low）
5. 如果数据不足以判断某项，明确标注"数据不足，需人工确认"

严格输出 JSON，不要输出其他内容。JSON 格式：
{
  "title": "模具设计评审报告",
  "overview": "一段 1-2 句话的整体概述",
  "sections": [
    {
      "heading": "关键问题",
      "type": "issues",
      "items": [
        {"level": "high", "content": "拔模角不足，建议增加到 2° 以上"},
        {"level": "medium", "content": "壁厚偏差较大，可能产生缩痕"}
      ]
    },
    {
      "heading": "工艺参数建议",
      "type": "params_table",
      "headers": ["参数", "推荐值", "说明"],
      "rows": [
        ["锁模力", "150 吨", "基于投影面积计算"],
        ["注塑温度", "230°C", "ABS 材料推荐值"]
      ]
    },
    {
      "heading": "改进建议",
      "type": "recommendations",
      "items": [
        "建议将主流道直径从 4mm 增加到 6mm",
        "建议在所有垂直面增加 1-2° 拔模角"
      ]
    }
  ],
  "overall_score": 7.5
}"""

_DESIGN_REPORT_USER = """请根据以下模具分析数据生成评审报告：

## 产品信息
- 文件：{filename}
- 材料：{material}
- 体积：{volume}
- 表面积：{surface_area}
- 边界框：{bbox}

## 检测特征
{features}

## 质量指标
{quality_metrics}

## 分模方案
{schemes}

## 制造参数
- 推荐模具材料：{mold_material}
- 推荐模具硬度：{mold_hardness}
- 预估锁模力：{clamping_force}
- 模具尺寸（长×宽×高）：{mold_size}
- 预估成型周期：{cycle_time}
- 拔模角：{draft_angle}
- 收缩率：{shrinkage_rate}

## 原始设计建议
{recommendations}

请生成 JSON 格式评审报告。issues 部分不要超过 8 条，每条内容简洁在一行内；
params_table 至少要包含锁模力、成型周期、模仁材料、推荐型腔数 4 行；
如果某项数据标记为"自动计算"或"自动选择"，请在说明中注明"需人工确认"；
overall_score 范围 1-10。"""

_SIDE_ACTION_ANALYSIS_SYSTEM = """你是一位资深注塑模具结构工程师。
请根据提供的 STP 分析结果，判断当前产品是否需要倒扣/抽芯机构，并输出标准化结论。

输出要求：
1. 严格输出 JSON，不要输出其他内容
2. 只允许基于已给数据判断，数据不足时必须标记为 manual_review
3. 结论面向工程评审，避免坐标、面索引、底层算法术语堆砌
4. 建议必须标准化、简洁、可执行

JSON 格式：
{
  "status": "required|not_required|manual_review",
  "confidence": 0.0,
  "conclusion": "一句中文结论",
  "mechanism_recommendation": "slider|lifter|mixed|none|manual_review",
  "summary": "一段 40-80 字中文摘要",
  "reasons": ["原因1", "原因2"],
  "standard_advice": ["建议1", "建议2"],
  "manual_review_items": ["复核项1", "复核项2"]
}"""

_SIDE_ACTION_ANALYSIS_USER = """请分析当前注塑件是否需要倒扣/抽芯机构：

## 产品信息
- 文件：{filename}
- 材料：{material}
- 边界框：{bbox}

## 特征检测
{features}

## 最优方案
{best_scheme}

## 规则分析结果
{side_actions}

## DFM 风险
{dfm_violations}

判断要求：
1. 如果规则结果明确显示无倒扣，可输出 not_required
2. 如果存在外侧倒扣，优先考虑 slider
3. 如果存在内侧倒扣，优先考虑 lifter
4. 如果内外侧倒扣同时存在，可输出 mixed
5. 如果数据不够支撑明确判断，输出 manual_review"""

_PARTING_SYSTEM = """你是一位注塑模具分模专家。
根据产品几何特征和多个候选分模方向的评分数据，推荐最优分模方向。

输出要求：严格输出 JSON，不要输出其他内容。
JSON 格式：
{
  "recommended_axis": "Z",
  "confidence": 0.85,
  "reasoning": "详细的中文推理过程...",
  "risk_notes": ["风险1", "风险2"],
  "rankings": [{"axis":"Z","rank":1,"score":92,"note":"..."}]
}"""

_PARTING_USER = """请评估以下候选分模方向并推荐最优方案：

产品几何：
- 边界框 (mm)：{bbox}
- 面法向分布：{normal_stats}
- 惯性矩：{inertia}

约束条件：
- 材料：{material}
- 型腔数：{cavity_count}
- 最大锁模力 (吨)：{max_clamping_force}
- 泡沫材料：{is_foam}

候选方案：
{schemes}

请综合评估制造可行性、成本和风险，给出推荐。"""


class LLMService:
    """LLM 增强分析服务（单例）"""

    def __init__(self):
        self._enabled = settings.LLM_ENABLED
        self._api_url = settings.LLM_API_URL.rstrip("/")
        self._api_key = settings.LLM_API_KEY
        self._model = settings.LLM_MODEL
        self._timeout = settings.LLM_TIMEOUT
        self._max_tokens = settings.LLM_MAX_TOKENS

        if self._enabled:
            logger.info(
                "LLM 增强分析已启用: model=%s endpoint=%s",
                self._model, self._api_url,
            )
        else:
            logger.info("LLM 增强分析未启用（设置 LLM_ENABLED=true 启用）")

    async def generate_design_report(
        self,
        analysis_result: Dict[str, Any],
        detailed_cavity_json: Optional[Dict[str, Any]] = None,
    ) -> Optional[Dict[str, Any]]:
        """生成模具设计评审报告 (结构化 JSON)"""
        if not self._enabled:
            return None

        try:
            prompt = self._build_design_report_prompt(analysis_result, detailed_cavity_json)
            response = await self._chat(
                _DESIGN_REPORT_SYSTEM,
                prompt,
                self._max_tokens,
                expect_json=True,
            )
            if not response:
                return None
            result = self._parse_json_response(response)
            if result:
                logger.info("LLM 设计报告生成成功 (%d sections)", len(result.get("sections", [])))
            return result
        except Exception as e:
            logger.warning("LLM 设计报告生成失败（不影响主流程）: %s", e)
            return None

    async def generate_side_action_analysis(
        self,
        analysis_result: Dict[str, Any],
        detailed_cavity_json: Optional[Dict[str, Any]] = None,
    ) -> Optional[Dict[str, Any]]:
        """生成倒扣/抽芯 AI 标准化分析"""
        if not self._enabled:
            return None

        try:
            prompt = self._build_side_action_prompt(analysis_result, detailed_cavity_json)
            response = await self._chat(
                _SIDE_ACTION_ANALYSIS_SYSTEM,
                prompt,
                min(self._max_tokens, 1200),
                expect_json=True,
            )
            if not response:
                return None
            result = self._parse_json_response(response)
            if result:
                logger.info(
                    "LLM 倒扣/抽芯分析生成成功: status=%s confidence=%s",
                    result.get("status"),
                    result.get("confidence"),
                )
            return result
        except Exception as e:
            logger.warning("LLM 倒扣/抽芯分析失败（不影响主流程）: %s", e)
            return None

    @staticmethod
    def compose_llm_report(
        design_report: Optional[Dict[str, Any]],
        side_action_analysis: Optional[Dict[str, Any]],
    ) -> Optional[str]:
        """将结构化报告和倒扣分析打包进 llm_report 字段，避免改动外部协议。

        设计报告以 <!--DESIGN_REPORT_BEGIN--> / <!--DESIGN_REPORT_END--> 包裹的 JSON 嵌入，
        倒扣分析以 <!--SIDE_ACTION_AI_BEGIN--> / <!--SIDE_ACTION_AI_END--> 包裹的 JSON 嵌入。
        """
        sections: List[str] = []
        if side_action_analysis:
            payload = json.dumps(side_action_analysis, ensure_ascii=False)
            sections.append(
                "<!--SIDE_ACTION_AI_BEGIN-->\n"
                f"{payload}\n"
                "<!--SIDE_ACTION_AI_END-->"
            )
        if design_report:
            payload = json.dumps(design_report, ensure_ascii=False)
            sections.append(
                "<!--DESIGN_REPORT_BEGIN-->\n"
                f"{payload}\n"
                "<!--DESIGN_REPORT_END-->"
            )
        merged = "\n\n".join(sections).strip()
        return merged or None

    async def recommend_parting_direction(
        self,
        geometry_data: Dict[str, Any],
        candidate_schemes: List[Dict[str, Any]],
        material: Dict[str, Any],
        cavity_count: int = 1,
    ) -> Optional[Dict[str, Any]]:
        """推荐最优分型方向"""
        if not self._enabled:
            return None

        try:
            prompt = self._build_parting_prompt(geometry_data, candidate_schemes, material, cavity_count)
            response = await self._chat(_PARTING_SYSTEM, prompt, min(self._max_tokens, 1200), expect_json=True)
            if response:
                result = self._parse_json_response(response)
                if result:
                    logger.info("LLM 分型推荐: %s (%.2f)", result.get("recommended_axis", "?"), result.get("confidence", 0))
                return result
            return None
        except Exception as e:
            logger.warning("LLM 分型推荐失败（不影响主流程）: %s", e)
            return None

    def _build_design_report_prompt(self, analysis_result, detailed_cavity_json) -> str:
        features = json.dumps(analysis_result.get("detected_features", []), ensure_ascii=False, indent=2)
        if len(features) > 4000:
            features = features[:4000] + "\n... (已截断)"

        schemes_text = ""
        if detailed_cavity_json:
            schemes = detailed_cavity_json.get("candidate_schemes", [])
            if schemes:
                schemes_text = json.dumps([{
                    "scheme_id": s.get("scheme_id"), "rank": s.get("rank"), "title": s.get("title"),
                    "score": s.get("score"), "summary": s.get("summary"),
                    "parting_axis": s.get("parting", {}).get("axis"),
                    "mold_structure_type": s.get("mold_structure_type"),
                    "dfm_violations": s.get("dfm_violations", []),
                } for s in schemes], ensure_ascii=False, indent=2)

        best = detailed_cavity_json.get("candidate_schemes", [{}])[0] if detailed_cavity_json else {}
        cd = best.get("cavity_data", {}) if isinstance(best, dict) else {}
        mfg = cd.get("manufacturing_info", {})
        meta = cd.get("metadata", {})

        return _DESIGN_REPORT_USER.format(
            filename=meta.get("file_name", "unknown.stp"),
            material=meta.get("selected_material", "ABS"),
            volume=f"{analysis_result.get('geometry_data', {}).get('volume', 0):.1f} mm³",
            surface_area=f"{analysis_result.get('geometry_data', {}).get('surface_area', 0):.1f} mm²",
            bbox=json.dumps(analysis_result.get("geometry_data", {}).get("bounding_box", {}), ensure_ascii=False),
            features=features or "无特征检测数据",
            quality_metrics=json.dumps(analysis_result.get("quality_metrics", {}), ensure_ascii=False, indent=2),
            schemes=schemes_text or "无分模方案数据",
            mold_material=mfg.get("mold_material", "自动选择"),
            mold_hardness=mfg.get("mold_hardness", "自动选择"),
            clamping_force=mfg.get("estimated_clamping_force", "自动计算"),
            mold_size=json.dumps(mfg.get("estimated_mold_size", {}), ensure_ascii=False),
            cycle_time=mfg.get("estimated_cycle_time", "自动计算"),
            draft_angle=f"{meta.get('draft_angle', 2.0)}°",
            shrinkage_rate="自动计算",
            recommendations=json.dumps(analysis_result.get("design_recommendations", []), ensure_ascii=False, indent=2) or "无",
        )

    def _build_side_action_prompt(self, analysis_result, detailed_cavity_json) -> str:
        features = json.dumps(
            analysis_result.get("detected_features", []),
            ensure_ascii=False,
            indent=2,
        )
        if len(features) > 2500:
            features = features[:2500] + "\n... (已截断)"

        best_scheme = {}
        if detailed_cavity_json:
            candidate_schemes = detailed_cavity_json.get("candidate_schemes", [])
            best_scheme_id = detailed_cavity_json.get("best_scheme_id")
            if candidate_schemes:
                best_scheme = candidate_schemes[0]
                if best_scheme_id:
                    for scheme in candidate_schemes:
                        if scheme.get("scheme_id") == best_scheme_id:
                            best_scheme = scheme
                            break

        cavity_data = best_scheme.get("cavity_data", {}) if isinstance(best_scheme, dict) else {}
        metadata = cavity_data.get("metadata", {}) if isinstance(cavity_data, dict) else {}
        side_actions = (
            best_scheme.get("side_actions")
            or cavity_data.get("side_actions")
            or {}
        )
        best_scheme_view = {
            "scheme_id": best_scheme.get("scheme_id"),
            "title": best_scheme.get("title"),
            "score": best_scheme.get("score"),
            "parting_axis": best_scheme.get("parting", {}).get("axis"),
            "mold_structure_type": best_scheme.get("mold_structure_type"),
            "undercut_regions_count": len(best_scheme.get("undercut_regions", []) or []),
        }
        side_actions_view = {
            "summary": side_actions.get("summary", {}),
            "recommendations": side_actions.get("recommendations", []),
            "slider_count": len(side_actions.get("slider_mechanisms", []) or []),
            "lifter_count": len(side_actions.get("lifter_mechanisms", []) or []),
        }
        dfm_violations = best_scheme.get("dfm_violations", []) if isinstance(best_scheme, dict) else []

        return _SIDE_ACTION_ANALYSIS_USER.format(
            filename=metadata.get("file_name", "unknown.stp"),
            material=metadata.get("selected_material", "ABS"),
            bbox=json.dumps(
                analysis_result.get("geometry_data", {}).get("bounding_box", {}),
                ensure_ascii=False,
            ),
            features=features or "无特征检测数据",
            best_scheme=json.dumps(best_scheme_view, ensure_ascii=False, indent=2),
            side_actions=json.dumps(side_actions_view, ensure_ascii=False, indent=2),
            dfm_violations=json.dumps(dfm_violations[:6], ensure_ascii=False, indent=2),
        )

    def _build_parting_prompt(self, geometry_data, candidate_schemes, material, cavity_count) -> str:
        bbox = geometry_data.get("bounding_box", {})
        axis_normal_stats = geometry_data.get("axis_normal_stats", {})
        inertia = geometry_data.get("inertia_matrix", [])
        inertia_diag = [inertia[i][i] if i < len(inertia) and i < len(inertia[i]) else 0.0 for i in range(3)]

        schemes_text = json.dumps([{
            "axis": s.get("parting", {}).get("axis") or s.get("axis"),
            "score": s.get("score"), "summary": s.get("summary"),
            "mold_structure_type": s.get("mold_structure_type"),
            "core_required": s.get("core_required"),
            "dfm_violations": s.get("dfm_violations", []),
            "undercut_regions_count": len(s.get("undercut_regions", [])),
            "score_breakdown": s.get("score_breakdown", {}),
        } for s in candidate_schemes], ensure_ascii=False, indent=2)

        return _PARTING_USER.format(
            bbox=json.dumps(bbox, ensure_ascii=False),
            normal_stats=json.dumps(axis_normal_stats, ensure_ascii=False),
            inertia=json.dumps(inertia_diag, ensure_ascii=False),
            material=material.get("name", "ABS"),
            cavity_count=cavity_count,
            max_clamping_force="3000 吨（最大）",
            is_foam="是" if material.get("is_foam") else "否",
            schemes=schemes_text,
        )

    async def _chat(self, system, user, max_tokens=2000, expect_json=False, temperature=0.3):
        url = f"{self._api_url}/chat/completions"
        headers = {"Authorization": f"Bearer {self._api_key}", "Content-Type": "application/json"}
        payload = {
            "model": self._model,
            "messages": [{"role": "system", "content": system}, {"role": "user", "content": user}],
            "max_tokens": max_tokens, "temperature": temperature,
        }
        if expect_json:
            payload["response_format"] = {"type": "json_object"}
        async with httpx.AsyncClient(timeout=self._timeout) as client:
            resp = await client.post(url, json=payload, headers=headers)
            resp.raise_for_status()
            content = resp.json()["choices"][0]["message"]["content"]
            return content.strip() if content else None

    @staticmethod
    def _parse_json_response(raw):
        try:
            return json.loads(raw)
        except json.JSONDecodeError:
            m = re.search(r"\{[\s\S]*\}", raw)
            if m:
                try:
                    return json.loads(m.group())
                except json.JSONDecodeError:
                    pass
        logger.warning("LLM JSON 解析失败: %s...", raw[:200])
        return None


llm_service = LLMService()