#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
将HTML文章转换为：
1. 公众号版本 - HTML文件（带样式，浏览器打开后可全选复制粘贴到微信公众号编辑器）
2. 知乎版本 - Markdown文件（可直接粘贴到知乎编辑器）
"""
from bs4 import BeautifulSoup
import re
HTML_FILE = 'insurance-guide/articles/v22-persona.html'
def parse_source():
with open(HTML_FILE, 'r', encoding='utf-8') as f:
html = f.read()
soup = BeautifulSoup(html, 'html.parser')
body = soup.find('div', class_='article-body')
return soup, body
def split_two_versions(body):
"""文章里包含【A】公众号版和【B】知乎版，按 h1 切分。"""
elements = list(body.children)
parts = {'wechat': [], 'zhihu': []}
current = None
for el in elements:
if getattr(el, 'name', None) is None:
continue
if el.name == 'h1':
text = el.get_text()
if 'A' in text and ('公众号' in text or '\u516c\u4f17\u53f7' in text):
current = 'wechat'
continue
if 'B' in text and ('知乎' in text or '\u77e5\u4e4e' in text):
current = 'zhihu'
continue
if '双版本' in text or '\u53cc\u7248\u672c' in text:
current = None
continue
if current:
parts[current].append(el)
return parts
def element_to_markdown(el):
"""将单个html元素转换为markdown字符串。"""
name = el.name
if name in ('h1',):
return f"# {el.get_text().strip()}\n\n"
if name == 'h2':
return f"## {el.get_text().strip()}\n\n"
if name == 'h3':
return f"### {el.get_text().strip()}\n\n"
if name == 'h4':
return f"#### {el.get_text().strip()}\n\n"
if name == 'hr':
return "\n---\n\n"
if name == 'blockquote':
lines = []
for child in el.find_all(['p'], recursive=False):
txt = inline_to_md(child).strip()
for line in txt.split('\n'):
lines.append(f"> {line}" if line.strip() else ">")
if not lines:
txt = inline_to_md(el).strip()
for line in txt.split('\n'):
lines.append(f"> {line}" if line.strip() else ">")
return "\n".join(lines) + "\n\n"
if name == 'p':
return inline_to_md(el).strip() + "\n\n"
if name == 'ul':
items = []
for li in el.find_all('li', recursive=False):
items.append(f"- {inline_to_md(li).strip()}")
return "\n".join(items) + "\n\n"
if name == 'ol':
items = []
for i, li in enumerate(el.find_all('li', recursive=False), 1):
items.append(f"{i}. {inline_to_md(li).strip()}")
return "\n".join(items) + "\n\n"
return ''
def inline_to_md(el):
"""将段落内联格式（strong/em/code）转换。"""
parts = []
for child in el.children:
if getattr(child, 'name', None) is None:
parts.append(str(child))
elif child.name == 'strong' or child.name == 'b':
parts.append(f"**{child.get_text()}**")
elif child.name == 'em' or child.name == 'i':
parts.append(f"*{child.get_text()}*")
elif child.name == 'code':
parts.append(f"`{child.get_text()}`")
elif child.name == 'br':
parts.append("\n")
else:
parts.append(child.get_text())
return ''.join(parts)
def to_markdown(elements):
out = []
for el in elements:
md = element_to_markdown(el)
if md:
out.append(md)
text = ''.join(out)
text = re.sub(r'\n{3,}', '\n\n', text)
return text.strip() + '\n'
# ---------------- 公众号 HTML ----------------
WECHAT_CSS = """
"""
def wechat_html(elements, title):
body_html = []
skip_next_blockquote = False
for el in elements:
# 把首个 blockquote（标题/副标题/配图建议块）剔除，因为它是给作者看的备注，不是正文
if el.name == 'blockquote' and not skip_next_blockquote:
skip_next_blockquote = True
continue
# 把内容里被错误识别为 h2 的 "---" 干掉
if el.name == 'h2' and el.get_text().strip() in ('---', '——'):
continue
body_html.append(str(el))
raw = ''.join(body_html)
return f"""
{title}
{WECHAT_CSS}
— CPA × 香港保险 —
在迅雷做了 8 年总账，
我为什么把简历交给了安盛
把上市公司的财务标准，搬到家庭资产负债表上
{raw}
"""
# ---------------- 主流程 ----------------
def main():
soup, body = parse_source()
parts = split_two_versions(body)
# 公众号 HTML
wechat_html_str = wechat_html(parts['wechat'],
'在迅雷做了8年总账，我为什么把简历交给了安盛')
with open('公众号_在迅雷做了8年总账.html', 'w', encoding='utf-8') as f:
f.write(wechat_html_str)
# 知乎 Markdown
zhihu_md = to_markdown(parts['zhihu'])
title = '一个8年迅雷总账CPA，为什么转身去做香港保险？'
zhihu_md = f"# {title}\n\n" + zhihu_md
with open('知乎_8年迅雷CPA转做香港保险.md', 'w', encoding='utf-8') as f:
f.write(zhihu_md)
# 同时给一份公众号 markdown 备用
wechat_md = to_markdown(parts['wechat'])
wechat_md = f"# 在迅雷做了8年总账，我为什么把简历交给了安盛\n\n" + wechat_md
with open('公众号_在迅雷做了8年总账.md', 'w', encoding='utf-8') as f:
f.write(wechat_md)
print('已生成：')
print(' 公众号_在迅雷做了8年总账.html （在浏览器打开 → 全选复制 → 粘贴到公众号编辑器，保留全部样式）')
print(' 公众号_在迅雷做了8年总账.md （Markdown版备用）')
print(' 知乎_8年迅雷CPA转做香港保险.md （直接粘贴到知乎编辑器，知乎原生支持Markdown）')
if __name__ == '__main__':
main()