diff_format_zxh.py 1.2 KB

12345678910111213141516171819202122232425262728293031
  1. # -*- coding: utf-8 -*-
  2. """比对 560822024-张学红.pdf 与 v5 提取器已知的章节结构"""
  3. import fitz, re, json
  4. doc = fitz.open(r'E:/cfc/docs/参考资料/560822024-张学红.pdf')
  5. full = []
  6. for i, page in enumerate(doc):
  7. t = page.get_text()
  8. full.append((i + 1, t))
  9. # 1. 收集所有页脚章节名(每页最后一行通常是章节名)
  10. print('===== 每页页脚章节名(去重按出现顺序)=====')
  11. seen = []
  12. for pno, t in full:
  13. lines = [l.strip() for l in t.split('\n') if l.strip()]
  14. if lines:
  15. foot = lines[-1]
  16. # 去掉 "N/169" 页码
  17. foot = re.sub(r'\s*\d+/169\s*$', '', foot).strip()
  18. if foot and (not seen or seen[-1][1] != foot):
  19. if not any(foot == s[1] for s in seen):
  20. seen.append((pno, foot))
  21. for pno, foot in seen:
  22. print(f' p{pno}: {foot}')
  23. # 2. 找疑似"指标详解/新指标"线索:雌激素、慢病控制
  24. print()
  25. print('===== 关键词命中 =====')
  26. for kw in ['雌激素', '慢病控制', 'GMHI', '健康总分', '肠道预测年龄', '肠型', '有益菌', '有害菌', '核心菌属', '菌群多样性']:
  27. pages = [p for p, t in full if kw in t]
  28. print(f' {kw}: {len(pages)} 页 -> {pages[:8]}')