文本数据清洗：去除纯数字、纯字母、手机号等

import pandas as pd
import re


# 获取所有句子，并且没有重复值
df1 = pd.read_csv("")
col1 = df1[["usermsg"]]
df2 = pd.read_csv("")
col2 = df2[["usermsg"]]
col = pd.concat([col1, col2])
col = col.drop_duplicates()
col = col["usermsg"].apply(lambda x: x.strip())


new_data = []
for i in col:
    pattern2 = "(?<!\d)(1\d{10})(?!\d)"
    phone_list2 = re.compile(pattern2).findall(i)
    if phone_list2:  # 剔除手机号
        pass
    elif "http" in i:  # 剔除网站
        pass
    elif i.isdigit():  # 剔除纯数字
        pass
    elif i.encode("UTF-8").isalpha():  # 剔除纯英文
        pass
    elif i.encode("UTF-8").isalnum():  # 剔除混杂的纯英文和数字
        pass
    elif "." in i:  # 剔除小数符号
        pass
    elif "/:" in i:  # 剔除表情符号
        pass
    elif "[" in i:  # 剔除表情符号
        pass
    elif "【" in i:  # 剔除特殊符号
        pass
    elif ":" in i:  # 剔除特殊符号
        pass
    elif "_" in i:
        pass
    elif "vx" in i:
        pass
    elif "wx" in i:
        pass
    elif "-" in i:
        pass
    elif "：" in i:
        pass
    else:
        new_data.append(i)


new_df = pd.DataFrame(data=new_data)
new_df = new_df.dropna()
new_df.to_csv("new_clean.csv")