guoju0808/ry-ocr/app/services/qr_decoder.py

"""电子发票二维码识别

国家税务总局规范的电子发票二维码内容格式 (8 字段逗号分隔):
    01,<type>,<invoice_code>,<invoice_no>,<amount>,<date>,<check_code>,<reserved>

例: 01,31,,24922000000006110014,39500.00,20240202,,A371
"""
from __future__ import annotations

from dataclasses import dataclass
from pathlib import Path
from typing import Optional, Tuple

import cv2
import numpy as np
from loguru import logger


@dataclass
class QRDecodeResult:
    """二维码识别结果"""
    invoice_no: Optional[str] = None      # 发票号码
    amount: Optional[float] = None        # 金额（小写）
    invoice_date: Optional[str] = None    # 开票日期 YYYY-MM-DD
    raw: str = ""                         # 二维码原始文本


def _try_detect(img: np.ndarray, qd: cv2.QRCodeDetector) -> str:
    """opencv 单图识别 + 多二维码识别, 任一成功即返回文本"""
    data, _, _ = qd.detectAndDecode(img)
    if data:
        return data
    try:
        retval, decoded_info, _, _ = qd.detectAndDecodeMulti(img)
        if decoded_info:
            # 取第一个非空的
            for d in decoded_info:
                if d:
                    return d
    except Exception:
        pass
    return ""


def _detect_qr(image_path: Path) -> str:
    """从图片里解二维码; 全图 + 四象限各扫一遍"""
    img = cv2.imread(str(image_path))
    if img is None:
        return ""

    qd = cv2.QRCodeDetector()
    # 1) 全图
    txt = _try_detect(img, qd)
    if txt:
        return txt

    # 2) 四象限 (二维码常在票面边角)
    h, w = img.shape[:2]
    for name, crop in [
        ("left-top",     img[: h // 2, : w // 2]),
        ("right-top",    img[: h // 2, w // 2 :]),
        ("left-bottom",  img[h // 2 :, : w // 2]),
        ("right-bottom", img[h // 2 :, w // 2 :]),
    ]:
        txt = _try_detect(crop, qd)
        if txt:
            logger.debug("QR found in {}", name)
            return txt

    # 3) 放大再试 (二维码像素过小的情况)
    scaled = cv2.resize(img, None, fx=2.0, fy=2.0, interpolation=cv2.INTER_CUBIC)
    return _try_detect(scaled, qd)


def _parse_qr_payload(raw: str) -> QRDecodeResult:
    """解析电子发票二维码内容 → QRDecodeResult

    格式不合规返回空对象 (字段全 None), 由调用方判定为「格式不正确」并 fallback。
    """
    if not raw:
        return QRDecodeResult(raw=raw or "")

    parts = raw.split(",")
    if len(parts) != 8:
        logger.debug("QR 字段数 {} != 8, 视为格式不正确", len(parts))
        return QRDecodeResult(raw=raw)

    result = QRDecodeResult(raw=raw)

    # parts[3] = 发票号
    invoice_no = parts[3].strip()
    if invoice_no and (10 <= len(invoice_no) <= 30):
        result.invoice_no = invoice_no

    # parts[4] = 金额
    amt_str = parts[4].strip()
    if amt_str:
        try:
            result.amount = round(float(amt_str), 2)
        except ValueError:
            pass

    # parts[5] = 开票日期 (YYYYMMDD)
    date_str = parts[5].strip()
    if len(date_str) == 8 and date_str.isdigit():
        result.invoice_date = f"{date_str[:4]}-{date_str[4:6]}-{date_str[6:8]}"

    return result


def decode_qr(image_path: Path) -> QRDecodeResult:
    """从图片文件解电子发票二维码

    返回 QRDecodeResult; 字段全 None 表示无二维码或格式不正确。
    调用方应据此判定 fallback。
    """
    try:
        raw = _detect_qr(image_path)
    except Exception as e:
        logger.warning("QR 检测异常 {}: {}", image_path.name, e)
        return QRDecodeResult()

    if not raw:
        return QRDecodeResult()

    parsed = _parse_qr_payload(raw)
    # 任一关键字段解出即视为「格式正确」
    if parsed.invoice_no or parsed.amount or parsed.invoice_date:
        logger.info("QR 解码成功: no={}, amt={}, date={}",
                    parsed.invoice_no, parsed.amount, parsed.invoice_date)
    else:
        logger.debug("QR 解出但字段无效: raw={}", raw[:80])
    return parsed


def has_any_field(qr: QRDecodeResult) -> bool:
    """QR 是否解出了至少一个核心字段"""
    return bool(qr.invoice_no or qr.amount or qr.invoice_date)