# This Python file uses the following encoding: utf-8 # @author runhey # github https://github.com/runhey import re import csv import pandas as pd def parse_one(value: str) -> tuple: if not value: print(f'value is None {value}') return None result = re.search(r"(.*)?——(.*)", value) if result: return result[1], result[2] else: print(f'parse error {value}') return None, None if __name__ == "__main__": file = 'data.xlsx' df_read = pd.read_excel(file, usecols='A') df_wirte = pd.DataFrame(columns=['question', 'answer']) print(df_read.columns) for index, row in df_read.iterrows(): question, answer = parse_one(row['question answering']) df_wirte.loc[index] = [question, answer] df_wirte.to_csv('data.csv', index=False, encoding='utf-8-sig')