# -*- coding: utf-8 -*-
"""
Created on Wed Dec 29 15:41:01 2021

@author: Sim
"""
from selenium import webdriver
from bs4 import BeautifulSoup
import pandas as pd

uri ="file:///D:/htex/Pythonbk/codesdata/webcrawl3.html"
driver = webdriver.Chrome('d:/python/webdrivers/chromedriver.exe')
driver.get(uri)
#----------------------------------------------
htmlsrc = driver.page_source
soup = BeautifulSoup(htmlsrc, 'html.parser')

lst1 = soup.select('html') # 언어
print(lst1[0]['lang'])

lst2 = soup.select('title')  # 제목 내용
print(lst2[0].text)
#-------------------------------------------------
lst3 = soup.select('#text')
for tag in lst3:
     print(tag.name)
#-------------------------------------------------
data = soup.select('pre')[0].text
data = data.split()
vname = data[0:3]  # 처음 세 개는 변수이름
data = data[3:]
id = tuple(map(int,data[0:len(data):3]))
x = tuple(map(int,data[1:len(data):3]))
y = tuple(map(int, data[2:len(data):3]))
df = pd.DataFrame({'id': id, 'response':x, 'indep': y})
print(df.corr())