"""接口协议 - Pydantic v2"""
from typing import List, Optional

from pydantic import BaseModel, Field


# ===== 单行 OCR 结果 =====

class OCRLine(BaseModel):
    """OCR 识别到的一行文字"""
    text: str = Field(..., description="识别文本")
    confidence: float = Field(..., ge=0, le=1, description="置信度 0~1")
    box: List[List[float]] = Field(default_factory=list, description="四点坐标 [[x1,y1],...]")


# ===== 发票字段 =====

class InvoiceFields(BaseModel):
    """结构化发票字段"""
    # 基本信息
    invoice_type: Optional[str] = Field(None, description="发票类型，如 增值税电子普通发票")
    invoice_no: Optional[str] = Field(None, description="发票号码")
    invoice_code: Optional[str] = Field(None, description="发票代码")
    invoice_date: Optional[str] = Field(None, description="开票日期 YYYY-MM-DD")

    # 金额
    amount: Optional[float] = Field(None, description="价税合计（小写）")
    amount_cn: Optional[str] = Field(None, description="价税合计（大写中文）")
    amount_pretax: Optional[float] = Field(None, description="不含税金额")
    tax_amount: Optional[float] = Field(None, description="税额")

    # 主体
    seller_name: Optional[str] = Field(None, description="销售方名称")
    seller_tax_no: Optional[str] = Field(None, description="销售方纳税人识别号")
    buyer_name: Optional[str] = Field(None, description="购买方名称")
    buyer_tax_no: Optional[str] = Field(None, description="购买方纳税人识别号")

    # 校验
    amount_match: Optional[bool] = Field(None, description="大写金额与小数金额是否一致")


# ===== 响应 =====

class InvoiceResult(BaseModel):
    """发票识别总响应"""
    success: bool = Field(..., description="是否识别成功")
    is_invoice: bool = Field(True, description="是否被判定为发票（false = 非发票图片）")
    raw_text: str = Field("", description="全部 OCR 文本拼接")
    lines: List[OCRLine] = Field(default_factory=list, description="分行识别结果")
    fields: InvoiceFields = Field(default_factory=InvoiceFields, description="抽取的结构化字段")
    page_count: int = Field(1, description="PDF 页数 / 图片=1")
    engine: str = Field("paddleocr", description="使用的 OCR 引擎")
    elapsed_ms: int = Field(0, description="识别耗时（毫秒）")
    error: Optional[str] = Field(None, description="失败原因")
    error_code: Optional[str] = Field(None, description="错误码: not_invoice / timeout / unsupported / ocr_failed")

    # QR 识别相关
    from_qr: bool = Field(False, description="是否从二维码取到了 3 个核心字段 (开票时间/发票号/金额)")
    qr_raw: Optional[str] = Field(None, description="二维码原始文本 (用于排查)")
    qr_error: Optional[str] = Field(None, description="二维码识别失败原因 (no_qr / bad_format)")


class HealthResponse(BaseModel):
    status: str = "ok"
    version: str
    engine_ready: bool


class PathRecognizeRequest(BaseModel):
    """按文件路径识别的请求体"""
    file_path: str = Field(..., description="服务器本地绝对路径", examples=["E:/invoice/abc.pdf"])