none_parser.py 779 B

123456789101112131415161718192021222324252627
  1. import sys, os
  2. sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
  3. from parser.document_parser import DocumentParser
  4. from logger.logging import setup_logging
  5. logger = setup_logging()
  6. class NoneParser(DocumentParser):
  7. """空解析器"""
  8. def parse(self, data: bytes) -> str:
  9. """返回空字符串"""
  10. logger.info(f"空解析器,不解析数据,默认返回空字符串")
  11. return ""
  12. def supports(self, content_type: str) -> bool:
  13. """空解析器,默认支持所有类型,不检查内容类型"""
  14. return True
  15. if __name__ == "__main__":
  16. parser = NoneParser()
  17. with open("../data/HR.pdf", "rb") as f:
  18. data = f.read()
  19. result = parser.parse(data)
  20. print(result[:200])