fix_images.py

#!/usr/bin/env python3
"""修复 HTML 中的图片"""

import re
import sys

html_file = "output/article_20260107_115239.html"

# 读取 HTML
with open(html_file, 'r', encoding='utf-8') as f:
    html = f.read()

# 图片路径
cover_image = "images/gemini_20260107_115148_1.png"
inline_images = [
    "images/gemini_20260107_115205_1.png",
    "images/gemini_20260107_115222_1.png",
    "images/gemini_20260107_115239_1.png"
]

# 1. 插入封面图
# 在标题前插入封面
title_pattern = r'(<h1>.*?</h1>)'
cover_html = f'<img src="{cover_image}" class="cover" alt="封面图">\n\n        '
html = re.sub(title_pattern, cover_html + r'\1', html, count=1)

# 2. 插入配图
# 在每个 h2 标签后插入图片
h2_pattern = r'(<h2>.*?</h2>)'
h2_tags = re.findall(h2_pattern, html)

for idx, h2 in enumerate(h2_tags[:len(inline_images)]):
    img_html = f'\n<img src="{inline_images[idx]}" class="inline-image">'
    html = html.replace(h2, h2 + img_html, 1)

# 保存
with open(html_file, 'w', encoding='utf-8') as f:
    f.write(html)

print(f"✅ 已修复 {html_file}")
print(f"   插入封面图: 1 张")
print(f"   插入配图: {len(inline_images)} 张")