# 安裝依賴
pip install paddlepaddle paddleocr openpyxl pillow opencv-python python-dotenv
# 郵件發(fā)送使用標(biāo)準(zhǔn)庫，無需額外安裝

from paddleocr import PaddleOCR

# 初始化OCR引擎（首次運(yùn)行會(huì)自動(dòng)下載模型）
ocr = PaddleOCR(
    use_angle_cls=True,   # 使用方向分類器，處理旋轉(zhuǎn)文字
    lang='ch',            # 中文模型
    show_log=False        # 關(guān)閉冗余日志
)

import re

def extract_invoice_info(ocr_result):
    """
    從PaddleOCR識(shí)別結(jié)果中提取發(fā)票關(guān)鍵信息
    ocr_result格式: list of [box, (text, confidence)]
    """
    full_text = ""
    for line in ocr_result:
        text = line[1][0]
        confidence = line[1][1]
        if confidence > 0.7:  # 只保留高置信度結(jié)果
            full_text += text + " "
    
    # 提取金額（支持多種寫法）
    amount_pattern = r'(?:金額|合計(jì)|總計(jì))[:：\s]*([￥￥]?(\d+(?:\.\d{1,2})?))'
    amount_match = re.search(amount_pattern, full_text)
    amount = amount_match.group(1) if amount_match else None
    
    # 提取開票日期（常見格式：YYYY年MM月DD日 或 YYYY-MM-DD）
    date_pattern = r'(?:開票日期|發(fā)票日期)[:：\s]*(\d{4}[年-]\d{1,2}[月-]\d{1,2}日?)'
    date_match = re.search(date_pattern, full_text)
    invoice_date = date_match.group(1) if date_match else None
    
    # 提取供應(yīng)商（銷售方名稱）
    seller_pattern = r'(?:銷售方|出售方|供應(yīng)商)[:：\s]*([\u4e00-\u9fa5a-zA-Z0-9（）()]+公司[\u4e00-\u9fa5]*)'
    seller_match = re.search(seller_pattern, full_text)
    seller = seller_match.group(1) if seller_match else None
    
    return {
        "amount": amount,
        "date": invoice_date,
        "supplier": seller,
        "raw_text": full_text
    }

def ocr_invoice_image(image_path):
    """對(duì)單張發(fā)票圖片執(zhí)行OCR并返回結(jié)構(gòu)化信息"""
    try:
        result = ocr.ocr(image_path, cls=True)
        if not result or not result[0]:
            raise ValueError("OCR未識(shí)別到任何文本")
        # result[0] 是圖片中所有文本行
        info = extract_invoice_info(result[0])
        return info
    except Exception as e:
        print(f"OCR識(shí)別失敗: {image_path}, 錯(cuò)誤: {e}")
        return None

pip install pdf2image
# 還需要安裝poppler（Windows需下載，Linux apt install poppler-utils）

from pdf2image import convert_from_path

def ocr_invoice_pdf(pdf_path):
    images = convert_from_path(pdf_path, dpi=200, first_page=1, last_page=1)
    if not images:
        return None
    # 將PIL Image轉(zhuǎn)換為臨時(shí)文件或直接處理
    import tempfile
    with tempfile.NamedTemporaryFile(suffix=".png", delete=False) as tmp:
        images[0].save(tmp.name, "PNG")
        result = ocr.ocr(tmp.name, cls=True)
        info = extract_invoice_info(result[0]) if result and result[0] else None
        return info

from openpyxl import load_workbook
import os

def append_to_excel(excel_path, invoice_info):
    """
    將發(fā)票信息追加到Excel文件的報(bào)銷明細(xì)表中
    假設(shè)工作表名為"報(bào)銷明細(xì)"，表頭為：日期, 供應(yīng)商, 金額, 備注
    """
    if not os.path.exists(excel_path):
        # 如果文件不存在，創(chuàng)建新文件并寫入表頭
        from openpyxl import Workbook
        wb = Workbook()
        ws = wb.active
        ws.title = "報(bào)銷明細(xì)"
        ws.append(["日期", "供應(yīng)商", "金額", "備注"])
    else:
        wb = load_workbook(excel_path)
        ws = wb["報(bào)銷明細(xì)"]
    
    # 追加一行數(shù)據(jù)
    row = [
        invoice_info.get("date", ""),
        invoice_info.get("supplier", ""),
        invoice_info.get("amount", ""),
        f"OCR自動(dòng)識(shí)別 {invoice_info.get('raw_text', '')[:20]}..."
    ]
    ws.append(row)
    
    # 保存文件
    wb.save(excel_path)
    print(f"已追加記錄到 {excel_path}")

import smtplib
from email.mime.multipart import MIMEMultipart
from email.mime.text import MIMEText
from email.mime.base import MIMEBase
from email import encoders
import os

def send_excel_by_email(receiver_email, excel_path, subject="報(bào)銷單待審核", body="請(qǐng)查收本周報(bào)銷明細(xì)"):
    """
    發(fā)送Excel文件作為附件到指定郵箱
    """
    # 郵箱配置（從環(huán)境變量讀取）
    smtp_server = os.getenv("SMTP_SERVER", "smtp.qq.com")
    smtp_port = int(os.getenv("SMTP_PORT", "465"))
    sender_email = os.getenv("SENDER_EMAIL")
    sender_password = os.getenv("SENDER_PASSWORD")  # 授權(quán)碼
    
    if not sender_email or not sender_password:
        raise ValueError("請(qǐng)?jiān)?env文件中配置發(fā)件郵箱和密碼")
    
    # 構(gòu)建郵件
    msg = MIMEMultipart()
    msg["From"] = sender_email
    msg["To"] = receiver_email
    msg["Subject"] = subject
    msg.attach(MIMEText(body, "plain", "utf-8"))
    
    # 添加附件
    with open(excel_path, "rb") as f:
        part = MIMEBase("application", "octet-stream")
        part.set_payload(f.read())
        encoders.encode_base64(part)
        filename = os.path.basename(excel_path)
        part.add_header("Content-Disposition", f"attachment; filename={filename}")
        msg.attach(part)
    
    # 發(fā)送
    with smtplib.SMTP_SSL(smtp_server, smtp_port) as server:
        server.login(sender_email, sender_password)
        server.sendmail(sender_email, receiver_email, msg.as_string())
    print(f"郵件已發(fā)送至 {receiver_email}")

SMTP_SERVER=smtp.qq.com
SMTP_PORT=465
SENDER_EMAIL=rpa@company.com
SENDER_PASSWORD=your_authorization_code
AUDITOR_EMAIL=auditor@company.com

from dotenv import load_dotenv
load_dotenv()

import os
import time
import shutil
from pathlib import Path
import logging
from dotenv import load_dotenv

from ocr_engine import ocr_invoice_image, ocr_invoice_pdf
from excel_handler import append_to_excel
from mail_sender import send_excel_by_email

# 加載配置
load_dotenv()
INPUT_DIR = "invoices"
PROCESSED_DIR = "processed"
OUTPUT_EXCEL = "output/報(bào)銷明細(xì).xlsx"
AUDITOR_EMAIL = os.getenv("AUDITOR_EMAIL")
SCAN_INTERVAL = 60  # 掃描間隔（秒）

# 配置日志
logging.basicConfig(
    level=logging.INFO,
    format="%(asctime)s - %(levelname)s - %(message)s",
    handlers=[
        logging.FileHandler("invoice_robot.log"),
        logging.StreamHandler()
    ]
)
logger = logging.getLogger(__name__)

def process_single_file(file_path):
    """處理單個(gè)發(fā)票文件"""
    logger.info(f"開始處理文件: {file_path}")
    ext = file_path.suffix.lower()
    
    # 1. OCR識(shí)別
    if ext in ['.jpg', '.jpeg', '.png']:
        info = ocr_invoice_image(str(file_path))
    elif ext == '.pdf':
        info = ocr_invoice_pdf(str(file_path))
    else:
        logger.warning(f"不支持的文件類型: {ext}")
        return False
    
    if not info:
        logger.error(f"OCR識(shí)別失敗: {file_path}")
        return False
    
    logger.info(f"識(shí)別結(jié)果: 金額={info['amount']}, 日期={info['date']}, 供應(yīng)商={info['supplier']}")
    
    # 2. 寫入Excel
    try:
        append_to_excel(OUTPUT_EXCEL, info)
    except Exception as e:
        logger.error(f"Excel寫入失敗: {e}")
        return False
    
    # 3. 移動(dòng)文件到已處理文件夾
    processed_path = Path(PROCESSED_DIR) / file_path.name
    shutil.move(str(file_path), str(processed_path))
    logger.info(f"文件已歸檔: {processed_path}")
    
    return True

def scan_and_process():
    """掃描文件夾，處理所有待處理文件"""
    input_path = Path(INPUT_DIR)
    if not input_path.exists():
        input_path.mkdir()
    
    files = list(input_path.glob("*.*"))
    if not files:
        logger.info("暫無待處理文件")
        return
    
    for file_path in files:
        if file_path.suffix.lower() in ['.jpg', '.jpeg', '.png', '.pdf']:
            process_single_file(file_path)
        else:
            logger.info(f"跳過非圖片/PDF文件: {file_path.name}")

def main():
    logger.info("智能票據(jù)處理機(jī)器人啟動(dòng)")
    logger.info(f"監(jiān)控文件夾: {INPUT_DIR}, 掃描間隔: {SCAN_INTERVAL}秒")
    
    # 確保輸出目錄存在
    Path("output").mkdir(exist_ok=True)
    Path(PROCESSED_DIR).mkdir(exist_ok=True)
    
    # 持續(xù)運(yùn)行模式（也可改為單次執(zhí)行后退出）
    try:
        while True:
            scan_and_process()
            time.sleep(SCAN_INTERVAL)
    except KeyboardInterrupt:
        logger.info("機(jī)器人已停止")

if __name__ == "__main__":
    main()

import cv2

def preprocess_image(image_path):
    img = cv2.imread(image_path, cv2.IMREAD_GRAYSCALE)
    # 自適應(yīng)閾值二值化
    img = cv2.adaptiveThreshold(img, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C, cv2.THRESH_BINARY, 11, 2)
    # 去噪
    img = cv2.medianBlur(img, 3)
    return img

from tenacity import retry, stop_after_attempt, wait_exponential

@retry(stop=stop_after_attempt(3), wait=wait_exponential(multiplier=1, min=4, max=10))
def send_excel_with_retry(*args, **kwargs):
    send_excel_by_email(*args, **kwargs)

FROM python:3.9-slim
RUN apt-get update && apt-get install -y poppler-utils
WORKDIR /app
COPY requirements.txt .
RUN pip install -r requirements.txt
COPY . .
CMD ["python", "main.py"]