wenbin5243 看看图片命名有没有规律，有就用迅雷批量下来，没有就找个爬虫脚本改一改

甘博摄影集，zlibrary直接搜，216MB，质量应该和网站上的差不多，我下载了几张，原图都是几百KB。

---

看了下，下载的epub里的图片是压缩过的。

下载TinyTask才35K录制一次操作，然后打开100个页面，自动重复操作。

录制速度慢无所谓，它能加速最高100倍

让Gemini写了个爬虫，试了下可以下载，用图片描述来命名下载的图片。

将下面的代码保存到 download.py

命令行中输入 Python3 download.py 

按回车下载，会自动创建文件夹。

（需要安装python3，命令行中输入 pip3 install requests,lxml 安装依赖库）

import requests
from lxml import html
import os
import time
import re
def download_images_with_descriptive_names():
    base_url = "?f%5Bcommon_model_name_ssi%5D%5B%5D=Item"
    download_folder = "gamble_descriptive_images"
    if not .exists(download_folder):
        os.makedirs(download_folder)
    total_pages = 250
    for page_num in range(1, total_pages + 1):
        page_url = f"{base_url}&page={page_num}"
        print(f"\n--- 正在处理第 {page_num} 页, URL: {page_url} ---")
        try:
            response = requests.get(page_url, timeout=10)
            response.raise_for_status()
        except  as e:
            print(f"获取第 {page_num} 页失败: {e}")
            continue
        tree = html.fromstring()
        image_blocks = tree.xpath('//div[contains(@class, "blacklight-image")]')
        if not image_blocks:
            print(f"第 {page_num} 页未找到任何图片信息块。")
            continue
        for i, block in enumerate(image_blocks, 1):
            thumbnail_link = block.xpath('.//img[contains(@class, "img-thumbnail")]/@src')
            if not thumbnail_link:
                continue
            thumbnail_link = thumbnail_link[0]
            
            image_description = block.xpath('.//a[@itemprop="name"]/text()')
            if not image_description:
                ark_id = thumbnail_link.split('/')[3]
                filename = ark_id + ".jpg"
            else:
                filename = image_description[0].strip()
                filename = re.sub(r'[\\/:*?"<>|]', '', filename)
                filename = filename[:100].strip() + ".jpg"
            if '!350,350' in thumbnail_link:
                high_res_link = thumbnail_link.replace('!350,350', 'full')
            else:
                continue
            full_download_url = f"}"
            file_path = .join(download_folder, filename)
            print(f"  > 正在处理第 {i} 个图片: {filename}")
            try:
                if not .exists(file_path):
                    image_data = requests.get(full_download_url, stream=True, timeout=30)
                    image_data.raise_for_status()
                    
                    with open(file_path, 'wb') as f:
                        for chunk in image_data.iter_content(chunk_size=8192):
                            f.write(chunk)
                    print(f"  > 下载成功: {filename}")
                else:
                    print(f"  > 文件已存在，跳过: {filename}")
            except  as e:
                print(f"  > 下载失败: {e}")
            time.sleep(0.5)
        print(f"\n第 {page_num} 页处理完毕。")
        time.sleep(2)
    print("\n所有图片处理完毕。")
if __name__ == "__main__":
    download_images_with_descriptive_names()

vDtv3vNZoE5d