帖子链接：求助部分日本图书馆以及大学馆藏照片下载方法 – 书格

使用工具：编程语言python、chrome浏览器、chatgpt以及kimi ai

说明：没有技术含量，仅做分享。

思路分析：首先我们要解析这个xml，获取关键信息。其次就是分析信息，构造url，下载碎图。最后就是按事先设定的图片命名方式，进行拼图。

测试链接1：https://opac.lib.takushoku-u.ac.jp/kyugaichi/htmls/resources/2013_013_001/root.xml

测试链接2：当麻練供養図 | 誕生寺所蔵

第一步：我们先把https://www.nara-wu.ac.jp/aic/gdb/mahoroba/y29/taimanerikuyouzu/limes/taimanerikuyouzu.html 改成

https://www.nara-wu.ac.jp/aic/gdb/mahoroba/y29/taimanerikuyouzu/resources/tanjoji_taimanerikuyouzu/root.xml

第二步使用pycharm或者其它能运行py脚本的ide工具运行下面代码即可。

缺点就是速度比较慢，比不上类似的软件，有些参数也可以修改。

import os
import requests
from PIL import Image
import xml.etree.ElementTree as ET
from datetime import datetime

timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
download_path = os.path.join('E:\\2025', f'download_{timestamp}')
os.makedirs(download_path, exist_ok=True)

# 下载和解析 XML
dzi = "https://opac.lib.takushoku-u.ac.jp/kyugaichi/htmls/resources/2013_013_001/root.xml"
response = requests.get(dzi)
xml_content = response.text

# 打印 XML 内容进行调试
print("XML 内容：")
print(xml_content)

root = ET.fromstring(xml_content)

# 获取 <image> 节点并提取信息 ，没有的可以换成resource
image_node = root.find('.//image')
if image_node is None:
image_node = root.find('.//resource')

if image_node is not None:
print("找到节点：", image_node.tag)
else:
print("未找到节点")

if image_node is not None:
wi = int(image_node.attrib['width'])
hi = int(image_node.attrib['height'])
tsize = int(image_node.attrib['tilewidth'])
else:
raise ValueError("XML 文件中未找到 <image> 节点")

# 输出获取的参数
print(f"Width: {wi}, Height: {hi}, Tile Width: {tsize}")

# 计算列数和行数
cols = (wi // tsize) + 1
rows = (hi // tsize) + 1

# 下载图像瓦片并拼接
if not os.path.exists(download_path):
os.makedirs(download_path)

for i in range(rows):
string = []
for j in range(cols):
nh = tsize if i < rows - 1 else hi % tsize or tsize # 确保至少有一个瓦片
nw = tsize if j < cols - 1 else wi % tsize or tsize

num1 = f"{j * tsize:05d}"
num2 = f"{i * tsize:05d}"
num3 = f"{nw:05d}"
num4 = f"{nh:05d}"
image_name = f"{num1}{num2}{num3}{num4}.jpg"
img_path = os.path.join(download_path, f"{j}_{i}.jpg")
file_url = f"{dzi.replace('root.xml', '0/')}{image_name}"

print(f"Downloading {file_url}")
img_response = requests.get(file_url)
if img_response.status_code == 200:
with open(img_path, "wb") as f:
f.write(img_response.content)
string.append(img_path)
else:
print(f"Failed to download {file_url}. Status code: {img_response.status_code}")

# 拼接当前行的图像
if string:
row_image = Image.open(string[0])
for img_path in string[1:]:
next_image = Image.open(img_path)
new_width = row_image.width + next_image.width
new_image = Image.new('RGB', (new_width, row_image.height))
new_image.paste(row_image, (0, 0))
new_image.paste(next_image, (row_image.width, 0))
row_image = new_image
row_image.save(os.path.join(download_path, f"row{i}.jpg"))

# 最后合并所有行的图像
final_images = [Image.open(os.path.join(download_path, f"row{i}.jpg")) for i in range(rows)]
full_image = Image.new('RGB', (wi, hi))
y_offset = 0
for img in final_images:
full_image.paste(img, (0, y_offset))
y_offset += img.height

full_image.save(os.path.join(download_path, "full.jpg"))
print(f"全图已保存为 {os.path.join(download_path, 'full.jpg')}")