# -*- coding: utf-8 -*-

"""
@Time    :  2023/2/27 18:24
@Author  : 
@FileName: 
@Software: 
@Describe:
"""
import pandas as pd
import difflib

path = "../data/11篇_yy.xlsx"
data = pd.read_excel(
    path
).values.tolist()


data_new = []
for i in data:
    data_1 = i[0]
    data_2 = i[1]
    str_sim_value = difflib.SequenceMatcher(None, data_1, data_2).quick_ratio()
    data_new.append(i + [str_sim_value])

data_new = sorted(data_new, key= lambda x:x[2], reverse=True)
df = pd.DataFrame(data_new)
df.to_excel("../data/11篇_yy_strsim.xlsx", index=None)