将语雀文档内容同步Hexo+Github

import sys
import requests
import json
import re
import os
import urllib.parse
from datetime import datetime

def save_page(book_id, slug):
    """下载文档内容（不替换图片链接）"""
    url = f'https://www.yuque.com/api/docs/{slug}?book_id={book_id}&merge_dynamic_data=false&mode=markdown'
    r = requests.get(url)
    if r.status_code != 200:
        print("文档下载失败:", slug, r.content)
        return None, None, None
    data = r.json().get('data', {})
    content = data.get('sourcecode', '')

    # 不再替换图片链接，保留原始链接

    created_at = data.get('created_at')
    published_at = data.get('published_at')
    updated_at = data.get('updated_at') or published_at or created_at
    return content, created_at, updated_at


def build_categories(doc, list_dict):
    """根据父子关系构建 categories 数组"""
    categories = []
    parent = doc.get('parent_uuid')
    while parent:
        parent_doc = list_dict.get(parent)
        if not parent_doc:
            break
        categories.insert(0, parent_doc.get('title', ''))
        parent = parent_doc.get('parent_uuid') or None
    return categories


def safe_datetime_str(iso_ts):
    """格式化 ISO 时间"""
    if not iso_ts:
        return ''
    try:
        if iso_ts.endswith('Z'):
            iso_ts = iso_ts[:-1]
        dt = datetime.fromisoformat(iso_ts)
        return dt.strftime('%Y-%m-%d %H:%M:%S')
    except Exception:
        return iso_ts


def get_book(url="https://www.yuque.com/076w/syst1m"):
    r = requests.get(url)
    m = re.findall(r"decodeURIComponent\(\"(.+)\"\)\);", r.text)
    if not m:
        print("未在页面中找到 window.appData 数据，请检查 URL")
        return
    docsjson = json.loads(urllib.parse.unquote(m[0]))

    book_id = str(docsjson['book']['id'])
    book_dir = f"download/{book_id}"
    os.makedirs(book_dir, exist_ok=True)

    list_dict = {doc['uuid']: doc for doc in docsjson['book']['toc']}

    for doc in docsjson['book']['toc']:
        if doc.get('type') == 'DOC' and doc.get('url'):
            content, created_at, updated_at = save_page(book_id, doc['url'])
            if content is None:
                continue

            categories = build_categories(doc, list_dict)

            fm_lines = ["---", f"title: {doc.get('title', '')}"]

            # 如果有上级目录，则写 categories 和 tags
            if categories:
                fm_lines.append(f"categories: {json.dumps(categories, ensure_ascii=False)}")
                fm_lines.append(f"tags: {json.dumps([categories[-1]], ensure_ascii=False)}")

            fm_lines.append(f"id: {doc.get('doc_id', '')}")
            fm_lines.append(f"date: {safe_datetime_str(created_at)}")
            fm_lines.append(f"updated: {safe_datetime_str(updated_at)}")
            fm_lines.append("---")
            fm = "\n".join(fm_lines) + "\n\n"

            safe_title = re.sub(r'[\\/*?:"<>|]', '_', doc.get('title', 'Untitled')).strip()
            md_file = os.path.join(book_dir, safe_title + '.md')

            with open(md_file, 'w', encoding='utf-8') as f:
                f.write(fm + content)

            print(f"已保存: {md_file}")


if __name__ == '__main__':
    if len(sys.argv) > 1:
        get_book(sys.argv[1])
    else:
        get_book()