"""
PMFlow AI 服务 - Pydantic 请求/响应模型

对应 ai_service 各核心函数的输入输出。
"""

from typing import List, Optional
from pydantic import BaseModel, Field


# ============================================================================
# 需求宪章
# ============================================================================


class CharterDraftResponse(BaseModel):
    """从信号生成的需求宪章草案"""

    title: str
    summary: str
    background: str
    objectives: List[str]
    scope: str
    success_metrics: List[str]
    risks: List[str]
    raw_content: Optional[str] = None


# ============================================================================
# 任务拆解
# ============================================================================


class TaskWithSkillRequirement(BaseModel):
    """带技能要求的任务项"""

    task_name: str
    description: str
    type: str  # frontend, backend, database, ai, devops
    estimated_hours: float
    required_skill_tag_ids: List[int] = Field(default_factory=list)
    required_skill_names: List[str] = Field(default_factory=list)
    priority: str = "medium"


class TaskDecompositionResponse(BaseModel):
    """宪章拆解任务响应"""

    tasks: List[TaskWithSkillRequirement]
    total_estimated_hours: float
    detected_skills: List[str]


# ============================================================================
# 人员推荐
# ============================================================================


class AssigneeRecommendation(BaseModel):
    """被推荐人"""

    user_id: str  # UUID 或 int 的字符串形式
    user_name: str
    match_score: float = Field(..., ge=0, le=100)
    reason: str
    current_workload: float
    proficiency_avg: float
    historical_performance: Optional[float] = None


class AssigneeRecommendationResponse(BaseModel):
    """人员推荐响应"""

    recommendations: List[AssigneeRecommendation]
    total_count: int


# ============================================================================
# 技能分数更新
# ============================================================================


class UserSkillScoreUpdate(BaseModel):
    """用户技能分数更新结果"""

    user_id: str
    skill_tag_id: str
    quality_score_delta: float
    efficiency_score_delta: float
    last_used_at: str


class UpdateSkillScoresResponse(BaseModel):
    """更新技能分数响应"""

    updated_count: int
    updates: List[UserSkillScoreUpdate]


# ============================================================================
# 认知对齐
# ============================================================================


class AlignmentScoreResponse(BaseModel):
    """认知对齐度评分响应"""

    user_id: str
    truth_node_id: str
    alignment_score: float = Field(..., ge=0, le=100)
    dimension_scores: Optional[dict] = None
    explanation: Optional[str] = None


# ============================================================================
# 认知冲突
# ============================================================================


class CognitiveConflict(BaseModel):
    """认知冲突项"""

    conflict_type: str  # e.g. definition, priority, scope
    description: str
    involved_user_ids: List[str]
    involved_user_names: List[str]
    severity: str  # low, medium, high


class CognitiveConflictsResponse(BaseModel):
    """认知冲突检测响应"""

    truth_node_id: str
    conflicts: List[CognitiveConflict]
    total_count: int


# ============================================================================
# 模型需求评估
# ============================================================================


class ModelRequirementEvaluation(BaseModel):
    """需求与模型匹配度评估"""

    compatibility_score: float = Field(..., ge=0, le=100)
    cost_estimate_usd: Optional[float] = None
    cost_estimate_tokens_per_day: Optional[int] = None
    risk_hints: List[str] = Field(default_factory=list)
    recommendations: List[str] = Field(default_factory=list)