Python批量处理PDF合同与发票的实战案例

import pandas as pd

# 模拟客户数据
data = {
    "客户编号": ["KH001", "KH002", "KH003"],
    "客户名称": ["北京科技有限公司", "上海信息有限公司", "广州贸易有限公司"],
    "合同金额": [150000, 98000, 210000],
    "签订日期": ["2026-07-01", "2026-07-05", "2026-07-10"],
    "项目内容": ["数据分析系统开发", "爬虫数据采集服务", "可视化平台搭建"],
}

df = pd.DataFrame(data)
df.to_excel("客户数据.xlsx", index=False)

pip install python-docx docx2pdf  # 先生成 Word，再转 PDF

from docx import Document
from docx.shared import Pt, Cm, Inches, RGBColor
from docx.enum.text import WD_ALIGN_PARAGRAPH
import pandas as pd
from datetime import datetime
import os

class ContractGenerator:
    """合同批量生成器"""

    def __init__(self, template_path):
        self.template_path = template_path

    def generate_contract(self, customer):
        """为单个客户生成合同"""
        doc = Document(self.template_path)

        # 替换占位符
        replacements = {
            "{{客户编号}}": customer["客户编号"],
            "{{客户名称}}": customer["客户名称"],
            "{{合同金额}}": f"{customer['合同金额']:,.0f}",
            "{{大写金额}}": self._num_to_cn(customer["合同金额"]),
            "{{签订日期}}": customer["签订日期"],
            "{{项目内容}}": customer["项目内容"],
        }

        for p in doc.paragraphs:
            for key, val in replacements.items():
                if key in p.text:
                    for run in p.runs:
                        if key in run.text:
                            run.text = run.text.replace(key, val)

        # 保存
        filename = f"合同_{customer['客户编号']}_{customer['客户名称']}.docx"
        filepath = os.path.join("生成合同", filename)
        os.makedirs("生成合同", exist_ok=True)
        doc.save(filepath)
        return filepath

    def _num_to_cn(self, num):
        """金额转大写（简化版）"""
        c_num = "零壹贰叁肆伍陆柒捌玖"
        c_unit = "分角元拾佰仟万拾佰仟亿"
        # 实际项目需要完善的大写金额转换
        return f"{num}元整"

    def batch_generate(self, excel_path):
        """批量生成"""
        df = pd.read_excel(excel_path)
        files = []
        for _, row in df.iterrows():
            path = self.generate_contract(row)
            files.append(path)
            print(f"已生成: {os.path.basename(path)}")
        return files

# 使用
generator = ContractGenerator("合同模板.docx")
files = generator.batch_generate("客户数据.xlsx")

from docx2pdf import convert

def batch_docx_to_pdf(docx_dir="生成合同", pdf_dir="生成PDF"):
    """批量把 Word 合同转成 PDF"""
    os.makedirs(pdf_dir, exist_ok=True)

    for f in os.listdir(docx_dir):
        if f.endswith(".docx"):
            docx_path = os.path.join(docx_dir, f)
            pdf_path = os.path.join(pdf_dir, f.replace(".docx", ".pdf"))
            convert(docx_path, pdf_path)
            print(f"已转换: {f} → {os.path.basename(pdf_path)}")

# 使用
batch_docx_to_pdf()

import pdfplumber
import pandas as pd
import re
import os

class InvoiceExtractor:
    """发票信息提取器"""

    def __init__(self):
        self.fields = ["发票号码", "开票日期", "金额", "税额", "价税合计",
                       "销售方名称", "购买方名称", "货物名称"]

    def extract_from_text(self, pdf_path):
        """从文字型 PDF 发票中提取信息"""
        info = {"文件": os.path.basename(pdf_path)}
        text = ""

        with pdfplumber.open(pdf_path) as pdf:
            for page in pdf.pages:
                text += page.extract_text() or ""

        # 发票号码
        match = re.search(r"发票号码[：:]\s*(\d+)", text)
        if match:
            info["发票号码"] = match.group(1)

        # 开票日期
        match = re.search(r"开票日期[：:]\s*(\d{4}[\-\年]\d{1,2}[\-\月]\d{1,2})", text)
        if match:
            info["开票日期"] = match.group(1).replace("年", "-").replace("月", "-")

        # 金额（价税合计）
        match = re.search(r"价税合计[（(]大写[）)][^0-9]*([\d,]+\.\d{2})", text)
        if match:
            info["价税合计"] = match.group(1).replace(",", "")

        # 销售方名称
        match = re.search(r"销售方[名称]?[：:]\s*(.+?)(?:\n|$)", text)
        if match:
            info["销售方名称"] = match.group(1).strip()

        # 购买方名称
        match = re.search(r"购买方[名称]?[：:]\s*(.+?)(?:\n|$)", text)
        if match:
            info["购买方名称"] = match.group(1).strip()

        return info

    def extract_from_table(self, pdf_path):
        """从发票 PDF 的表格中提取商品明细"""
        items = []
        with pdfplumber.open(pdf_path) as pdf:
            for page in pdf.pages:
                tables = page.extract_tables()
                for table in tables:
                    for row in table[1:]:  # 跳过表头
                        if row and row[0] and len(row) >= 4:
                            items.append({
                                "货物名称": row[0],
                                "数量": row[1] if len(row) > 1 else "",
                                "单价": row[2] if len(row) > 2 else "",
                                "金额": row[3] if len(row) > 3 else "",
                            })
        return items

    def batch_extract(self, pdf_dir):
        """批量提取"""
        all_data = []
        for f in os.listdir(pdf_dir):
            if f.endswith(".pdf"):
                info = self.extract_from_text(os.path.join(pdf_dir, f))
                all_data.append(info)
                print(f"已提取: {f}")

        df = pd.DataFrame(all_data)
        df.to_excel("发票信息汇总.xlsx", index=False)
        print(f"\n共提取 {len(all_data)} 张发票，已导出到 发票信息汇总.xlsx")
        return df

# 使用
extractor = InvoiceExtractor()
extractor.batch_extract("发票PDF")

import pytesseract
from PIL import Image
import pdfplumber
import io

def ocr_invoice(pdf_path):
    """OCR 识别扫描件发票"""
    with pdfplumber.open(pdf_path) as pdf:
        for i, page in enumerate(pdf.pages):
            # 转图片
            img = page.to_image(resolution=300)
            img_path = f"temp_page_{i}.png"
            img.save(img_path)

            # OCR 识别
            text = pytesseract.image_to_string(
                Image.open(img_path),
                lang="chi_sim+eng",  # 中文 + 英文
            )

            # 提取关键信息
            invoice_no = re.search(r"发票号码[：:]\s*(\d+)", text)
            total = re.search(r"价税合计[：:]?\s*[¥￥]?([\d,]+\.\d{2})", text)

            print(f"发票号码: {invoice_no.group(1) if invoice_no else '未识别'}")
            print(f"价税合计: {total.group(1) if total else '未识别'}")

# 使用（需要安装 Tesseract-OCR 引擎）
# ocr_invoice("扫描发票.pdf")

from PyPDF2 import PdfMerger, PdfReader, PdfWriter
import os

def merge_invoices(pdf_dir, output="合并发票.pdf"):
    """合并多个发票 PDF 为一个文件"""
    merger = PdfMerger()
    files = sorted([f for f in os.listdir(pdf_dir) if f.endswith(".pdf")])

    for f in files:
        merger.append(os.path.join(pdf_dir, f))

    merger.write(output)
    merger.close()
    print(f"合并完成，共 {len(files)} 个文件 → {output}")

def split_into_single_pages(pdf_path, output_dir="拆分明细"):
    """将合并的 PDF 按页拆分为单张"""
    os.makedirs(output_dir, exist_ok=True)
    reader = PdfReader(pdf_path)

    for i in range(len(reader.pages)):
        writer = PdfWriter()
        writer.add_page(reader.pages[i])
        output = os.path.join(output_dir, f"第{i+1}页.pdf")
        with open(output, "wb") as f:
            writer.write(f)

    print(f"已拆分 {len(reader.pages)} 页到 {output_dir}")

def daily_invoice_pipeline(input_dir):
    """每日发票处理流水线"""
    print("开始发票处理流水线...")

    # 1. 提取信息
    extractor = InvoiceExtractor()
    df = extractor.batch_extract(input_dir)

    # 2. 合并 PDF
    merge_invoices(input_dir, f"发票汇总_{datetime.now().strftime('%Y%m%d')}.pdf")

    # 3. 生成统计报表
    if not df.empty and "价税合计" in df.columns:
        total = df["价税合计"].sum()
        count = len(df)
        print(f"\n统计：共 {count} 张发票，总金额 {total:,.2f} 元")

    print("处理完成！")

# 使用
daily_invoice_pipeline("发票PDF")