"""
API调用日志服务层

提供API调用日志记录、计费和统计查询功能
需求: 10.1, 10.2, 10.3, 10.4, 10.5, 11.1, 11.2, 11.3, 11.4, 12.1, 12.2, 12.4
"""
import logging
import uuid
from datetime import datetime, date, timedelta
from decimal import Decimal
from typing import List, Optional, Tuple
from sqlalchemy.orm import Session
from sqlalchemy import func, desc, and_
from sqlalchemy.exc import IntegrityError

from app.models.api_call_log import ApiCallLog
from app.models.user import User
from app.models.platform_api_key import PlatformApiKey
from app.schemas.platform_stats import (
    StatsResponse,
    TrendItem,
    ModelDistItem,
    CallLogResponse
)

logger = logging.getLogger(__name__)


class ApiCallLogService:
    """API调用日志服务类"""

    def __init__(self, db: Session):
        self.db = db

    def create_log(
        self,
        user_id: str,
        api_key_id: int,
        model_id: Optional[int],
        model_name: str,
        is_local: bool,
        input_tokens: int,
        output_tokens: int,
        bill: Decimal,
        status: str = "success",
        error_message: Optional[str] = None,
        request_ip: Optional[str] = None
    ) -> ApiCallLog:
        """
        创建调用日志
        
        需求 10.3: 记录每次API调用的详细信息
        需求 10.4: 记录调用时间、模型、Token用量、费用
        
        Args:
            user_id: 用户ID
            api_key_id: API Key ID
            model_id: 模型ID
            model_name: 模型名称
            is_local: 是否为本地模型
            input_tokens: 输入Token数
            output_tokens: 输出Token数
            bill: 费用金额
            status: 调用状态
            error_message: 错误信息
            request_ip: 请求IP
            
        Returns:
            ApiCallLog: 创建的日志记录
        """
        log = ApiCallLog(
            user_id=user_id,
            api_key_id=api_key_id,
            model_id=model_id,
            model_name=model_name,
            is_local=is_local,
            input_tokens=input_tokens,
            output_tokens=output_tokens,
            bill=bill,
            status=status,
            error_message=error_message,
            request_ip=request_ip
        )
        self.db.add(log)
        self.db.commit()
        self.db.refresh(log)
        return log

    def get_user_stats(
        self,
        user_id: str,
        trend_days: int = 7,
        key_type: Optional[str] = None
    ) -> StatsResponse:
        """
        获取用户调用统计
        
        需求 11.1: 显示今日调用次数、本月调用次数、总调用次数
        需求 11.2: 显示今日消费金额、本月消费金额
        需求 11.3: 显示调用趋势图
        需求 11.4: 显示各模型调用占比饼图
        
        Args:
            user_id: 用户ID
            trend_days: 趋势数据天数
            key_type: 密钥类型: public 或 local
            
        Returns:
            StatsResponse: 统计响应
        """
        today = date.today()
        month_start = today.replace(day=1)
        
        # 构建基础查询条件
        base_filter = [ApiCallLog.user_id == user_id]
        if key_type:
            # 关联查询API Key类型
            base_filter.append(
                ApiCallLog.api_key_id.in_(
                    self.db.query(PlatformApiKey.id).filter(
                        PlatformApiKey.user_id == user_id,
                        PlatformApiKey.key_type == key_type
                    )
                )
            )
        
        # 今日统计
        today_stats = self.db.query(
            func.count(ApiCallLog.id).label('count'),
            func.coalesce(func.sum(ApiCallLog.bill), 0).label('cost')
        ).filter(
            *base_filter,
            func.date(ApiCallLog.created_at) == today
        ).first()
        
        # 本月统计
        month_stats = self.db.query(
            func.count(ApiCallLog.id).label('count'),
            func.coalesce(func.sum(ApiCallLog.bill), 0).label('cost')
        ).filter(
            *base_filter,
            func.date(ApiCallLog.created_at) >= month_start
        ).first()
        
        # 总计统计
        total_stats = self.db.query(
            func.count(ApiCallLog.id).label('count')
        ).filter(
            *base_filter
        ).first()
        
        # 趋势数据
        trend_start = today - timedelta(days=trend_days - 1)
        trend_data = self._get_trend_data(user_id, trend_start, today, key_type)
        
        # 模型分布
        model_distribution = self._get_model_distribution(user_id, key_type)
        
        return StatsResponse(
            today_calls=today_stats.count or 0,
            month_calls=month_stats.count or 0,
            total_calls=total_stats.count or 0,
            today_cost=Decimal(str(today_stats.cost or 0)),
            month_cost=Decimal(str(month_stats.cost or 0)),
            trend_data=trend_data,
            model_distribution=model_distribution
        )

    def _get_trend_data(
        self,
        user_id: str,
        start_date: date,
        end_date: date,
        key_type: Optional[str] = None
    ) -> List[TrendItem]:
        """获取趋势数据"""
        # 构建基础查询条件
        base_filter = [
            ApiCallLog.user_id == user_id,
            func.date(ApiCallLog.created_at) >= start_date,
            func.date(ApiCallLog.created_at) <= end_date
        ]
        if key_type:
            # 关联查询API Key类型
            base_filter.append(
                ApiCallLog.api_key_id.in_(
                    self.db.query(PlatformApiKey.id).filter(
                        PlatformApiKey.user_id == user_id,
                        PlatformApiKey.key_type == key_type
                    )
                )
            )
        
        results = self.db.query(
            func.date(ApiCallLog.created_at).label('date'),
            func.count(ApiCallLog.id).label('count'),
            func.coalesce(func.sum(ApiCallLog.bill), 0).label('cost')
        ).filter(
            *base_filter
        ).group_by(
            func.date(ApiCallLog.created_at)
        ).order_by(
            func.date(ApiCallLog.created_at)
        ).all()
        
        # 构建日期到数据的映射
        data_map = {str(r.date): (r.count, r.cost) for r in results}
        
        # 填充所有日期
        trend_data = []
        current = start_date
        while current <= end_date:
            date_str = str(current)
            count, cost = data_map.get(date_str, (0, 0))
            trend_data.append(TrendItem(
                date=date_str,
                count=count,
                cost=Decimal(str(cost))
            ))
            current += timedelta(days=1)
        
        return trend_data

    def _get_model_distribution(self, user_id: str, key_type: Optional[str] = None) -> List[ModelDistItem]:
        """获取模型分布数据"""
        # 构建基础查询条件
        base_filter = [ApiCallLog.user_id == user_id]
        if key_type:
            # 关联查询API Key类型
            base_filter.append(
                ApiCallLog.api_key_id.in_(
                    self.db.query(PlatformApiKey.id).filter(
                        PlatformApiKey.user_id == user_id,
                        PlatformApiKey.key_type == key_type
                    )
                )
            )
        
        results = self.db.query(
            ApiCallLog.model_name,
            func.count(ApiCallLog.id).label('count')
        ).filter(
            *base_filter
        ).group_by(
            ApiCallLog.model_name
        ).order_by(
            desc(func.count(ApiCallLog.id))
        ).limit(10).all()
        
        total = sum(r.count for r in results)
        if total == 0:
            return []
        
        return [
            ModelDistItem(
                model_name=r.model_name,
                count=r.count,
                percentage=round(r.count / total * 100, 2)
            )
            for r in results
        ]

    def get_call_logs(
        self,
        user_id: str,
        start_date: Optional[date] = None,
        end_date: Optional[date] = None,
        model_id: Optional[int] = None,
        api_key_id: Optional[int] = None,
        key_type: Optional[str] = None,
        page: int = 1,
        page_size: int = 20
    ) -> Tuple[List[CallLogResponse], int]:
        """
        获取调用日志列表
        
        需求 12.1: 支持按时间范围、模型、API Key进行筛选
        需求 12.2: 显示调用时间、模型名称、Token用量、费用金额、使用的API Key
        需求 12.4: 支持分页查询
        
        Args:
            user_id: 用户ID
            start_date: 开始日期
            end_date: 结束日期
            model_id: 模型ID
            api_key_id: API Key ID
            key_type: 密钥类型: public 或 local
            page: 页码
            page_size: 每页数量
            
        Returns:
            Tuple[List[CallLogResponse], int]: (日志列表, 总数)
        """
        query = self.db.query(ApiCallLog).filter(ApiCallLog.user_id == user_id)
        
        # 应用筛选条件
        if start_date:
            query = query.filter(func.date(ApiCallLog.created_at) >= start_date)
        if end_date:
            query = query.filter(func.date(ApiCallLog.created_at) <= end_date)
        if model_id:
            query = query.filter(ApiCallLog.model_id == model_id)
        if api_key_id:
            query = query.filter(ApiCallLog.api_key_id == api_key_id)
        if key_type:
            # 关联查询API Key类型
            query = query.filter(
                ApiCallLog.api_key_id.in_(
                    self.db.query(PlatformApiKey.id).filter(
                        PlatformApiKey.user_id == user_id,
                        PlatformApiKey.key_type == key_type
                    )
                )
            )
        
        # 获取总数
        total = query.count()
        
        # 分页查询
        logs = query.order_by(desc(ApiCallLog.created_at)).offset(
            (page - 1) * page_size
        ).limit(page_size).all()
        
        # 获取API Key前缀映射
        api_key_ids = [log.api_key_id for log in logs if log.api_key_id]
        api_key_map = {}
        if api_key_ids:
            api_keys = self.db.query(PlatformApiKey).filter(
                PlatformApiKey.id.in_(api_key_ids)
            ).all()
            api_key_map = {k.id: k.api_key_prefix for k in api_keys}
        
        from app.models.model import ModelNew, ModelPriceNew
        from sqlalchemy.orm import selectinload
        model_ids = [log.model_id for log in logs if log.model_id]
        price_map = {}
        if model_ids:
            # 用 selectinload 一次性加载价格，消灭 N+1
            models = self.db.query(ModelNew).options(
                selectinload(ModelNew.prices)
            ).filter(ModelNew.id.in_(model_ids)).all()
            for m in models:
                active_price = next((p for p in (m.prices or []) if p.is_active), None)
                if active_price:
                    price_map[m.id] = active_price
        
        # 构建响应
        result = []
        for log in logs:
            price_info = price_map.get(log.model_id) if log.model_id else None
            result.append(CallLogResponse(
                id=log.id,
                model_name=log.model_name,
                is_local=log.is_local,
                input_tokens=log.input_tokens,
                output_tokens=log.output_tokens,
                bill=Decimal(str(log.bill)),
                status=log.status,
                api_key_prefix=api_key_map.get(log.api_key_id, "N/A"),
                created_at=log.created_at,
                input_price=Decimal(str(price_info.input_price_discounted)) if price_info else None,
                output_price=Decimal(str(price_info.output_price_discounted)) if price_info else None
            ))
        
        return result, total