parent
ee10051eb6
commit
c1cf26483b
@ -1,2 +1,28 @@
|
||||
# python
|
||||
|
||||
import requests
|
||||
from bs4 import BeautifulSoup
|
||||
url = 'http://58921.com/alltime?page='
|
||||
headers = {
|
||||
'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/70.0.3538.25 Safari/537.36 Core/1.70.3776.400 QQBrowser/10.6.4212.400'
|
||||
}
|
||||
aulist=[]
|
||||
def movie_get():
|
||||
for i in range(2):
|
||||
r = requests.get(url=url+str(i), headers=headers)
|
||||
r.encoding = 'utf-8'
|
||||
soup=BeautifulSoup(r.text,"lxml")
|
||||
ltr=soup.find_all("tr")
|
||||
for tr in ltr:
|
||||
ltd=tr.find_all("td")
|
||||
if len(ltd)==0:
|
||||
continue
|
||||
sli=[]
|
||||
for i in [1,2,6]:
|
||||
sli.append(ltd[i].text)
|
||||
aulist.append(sli)
|
||||
def movie_write():
|
||||
with open('movies.txt','w') as f1:
|
||||
for x in aulist:
|
||||
f1.write(x[0]+'|'+x[1]+'|'+x[2]+'\n')
|
||||
movie_get()
|
||||
movie_write()
|
||||
|
||||
Loading…
Reference in new issue