preprocessing_test.py 385 B

1234567891011121314151617
  1. import codecs
  2. import pandas as pd
  3. from bs4 import BeautifulSoup
  4. from BiddingKG.dl.interface.extract import predict
  5. def test():
  6. df = pd.read_excel("has_table_no_attach.xlsx")
  7. for index, row in df.iterrows():
  8. if index % 100 == 0:
  9. print("Loop", index)
  10. text = row['dochtmlcon']
  11. predict(str(index), text)
  12. if __name__ == "__main__":
  13. test()