添加pdf的目录

 

1.安装这个包，

PyPDF2>=3.0.0

#!/usr/bin/env python
# -*- coding: utf-8 -*-
"""
PDF目录添加工具
从txt文件读取目录信息并添加到PDF文件中
"""


import re
import sys
import glob
import os
from PyPDF2 import PdfReader, PdfWriter




def parse_toc_from_txt(txt_file):
    """
    从txt文件解析目录信息
    """
    toc_items = []



    with open(txt_file, 'r', encoding='utf-8') as f:
        lines = f.readlines()



    for line in lines:
        line = line.strip()
        if not line or line == '目录正文':
            continue



        # 匹配格式：章节名称 + 点号 + 页码
        match = re.match(r'(.+?)\s*[\.。]+\s*(\d+)\s*$', line)
        if match:
            title = match.group(1).strip()
            page_num = int(match.group(2))
            toc_items.append((title, page_num))



    return toc_items




# 起始页偏移量（第一章从第8页开始，目录中写的是第1页，所以偏移量为7）
START_PAGE_OFFSET = 7




def find_pdf_file():
    """自动查找当前目录的PDF文件（排除output文件）"""
    pdf_files = glob.glob('*.pdf')
    # 排除output文件
    pdf_files = [f for f in pdf_files if 'output' not in f.lower()]
    return pdf_files[0] if pdf_files else None




def add_bookmarks_to_pdf(pdf_file, output_file, toc_items):
    """
    为PDF添加书签/目录
    """
    reader = PdfReader(pdf_file)
    writer = PdfWriter()



    # 复制所有页面
    for page in reader.pages:
        writer.add_page(page)



    # 添加书签
    for title, page_num in toc_items:
        # PDF页码从0开始，加上起始页偏移量
        pdf_page_index = page_num - 1 + START_PAGE_OFFSET



        if pdf_page_index < 0:
            pdf_page_index = 0
        elif pdf_page_index >= len(reader.pages):
            pdf_page_index = len(reader.pages) - 1



        writer.add_outline_item(title, pdf_page_index)



    # 保存输出文件
    with open(output_file, 'wb') as f:
        writer.write(f)



    print(f"成功添加 {len(toc_items)} 个目录项")
    print(f"输出文件: {output_file}")




def main():
    """主函数"""
    # 默认文件名
    txt_file = '目录.txt'
    output_file = 'output_with_toc.pdf'



    # 自动查找PDF文件
    pdf_file = find_pdf_file()
    if not pdf_file:
        print("错误: 当前目录未找到PDF文件")
        return



    # 命令行参数支持
    if len(sys.argv) >= 2:
        txt_file = sys.argv[1]
    if len(sys.argv) >= 3:
        output_file = sys.argv[2]



    print(f"读取目录文件: {txt_file}")
    print(f"读取PDF文件: {pdf_file}")
    print(f"输出文件: {output_file}")
    print("-" * 50)



    # 解析目录
    toc_items = parse_toc_from_txt(txt_file)



    if not toc_items:
        print("警告: 未找到任何目录项")
        return



    print(f"找到 {len(toc_items)} 个目录项:")
    for i, (title, page) in enumerate(toc_items, 1):
        print(f"  {i}. {title} - 第{page}页")
    print("-" * 50)



    # 添加书签到PDF
    add_bookmarks_to_pdf(pdf_file, output_file, toc_items)




if __name__ == '__main__':
    main()