在python中清理文本文件并导入为pandas dataframe

import pandas as pd
import re

with open('file.txt', 'r') as file:
    data = file.read()

cleaned_data = re.sub(r'[^a-zA-Z0-9\s]', '', data)  # 去除非字母数字字符
cleaned_data = re.sub(r'\s+', ' ', cleaned_data)  # 去除多余的空格
cleaned_data = cleaned_data.lower()  # 转换为小写字母

df = pd.DataFrame(cleaned_data.split(), columns=['text'])

import pandas as pd
import re

with open('file.txt', 'r') as file:
    data = file.read()

cleaned_data = re.sub(r'[^a-zA-Z0-9\s]', '', data)
cleaned_data = re.sub(r'\s+', ' ', cleaned_data)
cleaned_data = cleaned_data.lower()

df = pd.DataFrame(cleaned_data.split(), columns=['text'])