import requests
from bs4 import BeautifulSoup

movielinkArray = []
titleArray = []
combineArray = []
reviewsArray=[]
movielink2Array = []
idArray=[]
id=''
link=''
def start():
    for id in range(11):
        link='https://tw.movies.yahoo.com/movie_intheaters.html?p='+str(id+1)
        res = requests.get(link)
        soup = BeautifulSoup(res.text, 'html.parser')
        print("第"+str(id+1)+"頁========================================")
        for link in soup.find_all('div',class_="row-container"):
            titleArray.append(link.find('h4').text)
            movielinkArray.append(link.find('h4').find('a').get('href'))
            #print(link.find('h4').find('a').get('href'))
            firstlevel(movielinkArray)




def firstlevel(movielinkArray):
     link=''.join(movielinkArray)
     res = requests.get(link)
     soup = BeautifulSoup(res.text, 'html.parser')
     for link in soup.find_all('li',class_="last"):
         print("https://tw.movies.yahoo.com"+link.find('span').find('a').get('href'))
         movielink2Array.append("https://tw.movies.yahoo.com"+link.find('span').find('a').get('href'))


if __name__=='__main__':
    start()